THE NEAR-SURFACE LAYER OF THE OCEAN
taller, more upright parts of streaks. Well-formed horseshoe vortices can
then assume a life of their own, continuing to grow by taking vorticity and
turbulent kinetic energy from the mean flow itself. The mean shear also
rotates these coherent vortices forward until, by a combination of growth and
rotation, they contact the ground to form a dam with strong inflows along the
ground produced by the rotation of the vortex arms and the main flow
presenting pressing in from behind. With nowhere else to go, the trapped air
squirts backwards and outwards into the flow. This squirt is usually
described as an ejection/sweep event. Figure 5-50 illustrates a schematic of
the first ejection formation on a streak.
McNaughton and Brunet (2002) assume that their model is consistent
with information from many sources and provide a historical excursion into
studies of the bursting process. Theodorsen (1952) was probably the first
who gave theoretical arguments for initiation of horseshoe vortices about
slow-moving masses of fluid attached to the ground, with subsequent vortex
roll-up and ejection. In the 1950’s, he could not know of wall streaks nor
appreciate the power of the ejection. In a series of laboratory experiments,
Kline et al. (1967) discovered wall streaks and described their oscillation and
break-up with sudden ejection of fluid from very near the wall. Kline et al.
(1967) did not detect the overtaking fluid or the horseshoe vortices. Corino
and Brodkey (1969) observed the colliding masses of fluid and realized the
power of the ejections in the boundary-layer dynamics; their visualization
methods unfortunately could not detect the formation of horseshoe vortices.
Hinze (1975) and Offen and Kline (1975) proposed a relationship between
wall streaks and horseshoe vortices. Hagen and Kurosaka (1993) were the
first who demonstrated the connection between horseshoe vortices and
powerful ejections. From a large-eddy simulation (LES) of a convective
atmospheric boundary layer, Lin (2000) deduced a sequence of horseshoe
eddies followed by a vigorous ejection. Finally, McNaughton and Brunet
(2002) used his results to develop the schematic diagram shown in Figure
5-50.
McNaughton and Brunet’s mechanism thus appears to be consistent with
a wide range of results from laboratory and atmospheric boundary layer
experiments. The new element introduced by these researchers is the
pressure mechanism for streak formation. This mechanism has an important
consequence, since it allows the wall streaks to form within fully turbulent
layers. Previously, wall streak formation has been associated only with
viscous sublayers, though the importance of pressure rather than vorticity in
creating motions near the ground was shown from LES results by Moeng
(see Peltier et al., 1996).
366
taller, more upright parts of streaks. Well-formed horseshoe vortices can
then assume a life of their own, continuing to grow by taking vorticity and
turbulent kinetic energy from the mean flow itself. The mean shear also
rotates these coherent vortices forward until, by a combination of growth and
rotation, they contact the ground to form a dam with strong inflows along the
ground produced by the rotation of the vortex arms and the main flow
presenting pressing in from behind. With nowhere else to go, the trapped air
squirts backwards and outwards into the flow. This squirt is usually
described as an ejection/sweep event. Figure 5-50 illustrates a schematic of
the first ejection formation on a streak.
McNaughton and Brunet (2002) assume that their model is consistent
with information from many sources and provide a historical excursion into
studies of the bursting process. Theodorsen (1952) was probably the first
who gave theoretical arguments for initiation of horseshoe vortices about
slow-moving masses of fluid attached to the ground, with subsequent vortex
roll-up and ejection. In the 1950’s, he could not know of wall streaks nor
appreciate the power of the ejection. In a series of laboratory experiments,
Kline et al. (1967) discovered wall streaks and described their oscillation and
break-up with sudden ejection of fluid from very near the wall. Kline et al.
(1967) did not detect the overtaking fluid or the horseshoe vortices. Corino
and Brodkey (1969) observed the colliding masses of fluid and realized the
power of the ejections in the boundary-layer dynamics; their visualization
methods unfortunately could not detect the formation of horseshoe vortices.
Hinze (1975) and Offen and Kline (1975) proposed a relationship between
wall streaks and horseshoe vortices. Hagen and Kurosaka (1993) were the
first who demonstrated the connection between horseshoe vortices and
powerful ejections. From a large-eddy simulation (LES) of a convective
atmospheric boundary layer, Lin (2000) deduced a sequence of horseshoe
eddies followed by a vigorous ejection. Finally, McNaughton and Brunet
(2002) used his results to develop the schematic diagram shown in Figure
5-50.
McNaughton and Brunet’s mechanism thus appears to be consistent with
a wide range of results from laboratory and atmospheric boundary layer
experiments. The new element introduced by these researchers is the
pressure mechanism for streak formation. This mechanism has an important
consequence, since it allows the wall streaks to form within fully turbulent
layers. Previously, wall streak formation has been associated only with
viscous sublayers, though the importance of pressure rather than vorticity in
creating motions near the ground was shown from LES results by Moeng
(see Peltier et al., 1996).
366