A Measure of Scale Dependent Asymmetry in Turbulent Boundary Layer Flows Scaling and Reynolds Number Similarity

J. Fluid Mech. (2016), vol. 797, pp. 549–563.
c Cambridge University Press 2016 549

doi:10.1017/jfm.2016.294
A measure of scale-dependent asymmetry in

turbulent boundary layer flows: scaling and
Reynolds number similarity
Arvind Singh1, †, Kevin B. Howard2 and Michele Guala2

1 Department of Civil, Environmental and Construction Engineering, University of Central Florida,
Orlando, FL 32816, USA
2 St. Anthony Falls Laboratory, Department of Civil, Environmental and Geo-Engineering,
University of Minnesota, Minneapolis, MN 55414, USA
(Received 19 September 2015; revised 18 April 2016; accepted 19 April 2016;

first published online 24 May 2016)
The distribution of temporal scale-dependent streamwise velocity increments is

investigated in turbulent boundary layer flows at laboratory and atmospheric Reynolds
numbers, using the St. Anthony Falls Laboratory wind tunnel and the Surface Layer
Turbulence and Environmental Science Test dataset, respectively. The third-order
moments of velocity increments, or asymmetry index A(a, z), is computed for varying
wall distance z and time scale separation a, where it was observed to leave a robust,
distinct signature in the form of a hump, independent of Reynolds number and located
across the inertial range. The hump is observed in wall region limited to z+ < 5 × 103 ,
with a tendency to shift towards smaller time scales as the surface is approached
(z+ < 70). Comparing the two datasets, the hump, and its location, are found to obey
inner wall scaling and is regarded as a genuine feature of the canonical turbulent
boundary layer. The magnitude cumulant analysis of the scale-dependent velocity
increments further reveals that intermittency is also enhanced near the wall, in the
same flow region where the asymmetry signature was observed. The combination
of asymmetry and intermittency is inferred to point at non-local energy transfer and
scale coupling across a range of scales. From a turbulent structure perspective, such
non-local energy transfer can be seen as the result of strong scale-interaction processes
between outer scale motions in the logarithmic layer impacting and distorting smaller
scales at the wall, through abrupt energy transfer across scales bypassing the typical
energy cascade of the inertial range.
Key words: atmospheric surface layer, intermittency, wall turbulence
1. Introduction
Scale interactions have been extensively addressed in canonical zero-pressure-
gradient turbulent boundary layer flows over a wide range of smooth and rough
surfaces, or densely distributed roughness elements (see Warhaft 2002; Poggi et al.
2004; Hutchins & Marusic 2007a; Guala, Metzger & McKeon 2011, among others).
† Email address for correspondence: Arvind.Singh@ucf.edu
Downloaded from http:/www.cambridge.org/core. Indian Institute of Technology Guwahati, on 04 Nov 2016 at 11:06:43, subject to the Cambridge Core terms of use,
available at http:/www.cambridge.org/core/terms. http://dx.doi.org/10.1017/jfm.2016.294
550 A. Singh, K. B. Howard and M. Guala
This includes the smooth-wall turbulent boundary layer where the largest scales such
as very-large-scale motions (VLSMs) and superstructures (Kim & Adrian 1999; Guala,
Hommema & Adrian 2006; Hutchins & Marusic 2007b; Smits, McKeon & Marusic
2011) have been observed to interact with the small-scale turbulence populating
the near-wall region (Hutchins & Marusic 2007a; Mathis, Hutchins & Marusic 2009;
Marusic, Mathis & Hutchins 2010). The result of such interactions has been quantified
in terms of amplitude modulation, anisotropy and inhomogeneity of the dissipative
scales, manifested through an increased intermittency (Laval, Dubrulle & Nazarenko
2001; Mininni, Alexakis & Pouquet 2006; Guala, Metzger & McKeon 2010; de
Silva et al. 2015), in particular, near the wall. In such a region, strong coupling
between widely separated scales in the spectral domain results in a ‘non-local’ energy
transfer (Yeung, Brasseur & Wang 1995; Warhaft 2002; Poggi et al. 2004; Mininni
et al. 2006; Keylock, Singh & Foufoula-Georgiou 2013; Keylock et al. 2014; Singh,
Howard & Guala 2014; de Silva et al. 2015). This mechanism is different from the
‘local’ energy transfer between neighbouring scales in a Fourier sense, hypothesized
in the Kolmogorov (1941) model of energy cascade, under the small-scale isotropy
assumption. It is argued that, when turbulence does not have the time/space to
dissipate energy at the rate required by the energy-containing eddies, non-uniform
transfer of energy takes place from large to smaller scales, resulting in abrupt intense
gradients in the velocity time series. These intense events are amplified in high-order
structure function statistics causing a deviation from the Kolmogorov (1941) prediction
(Menevau & Sreenivasan 1991; Benzi et al. 1993; Guala et al. 2010). In addition,
the same rare, highly dissipative events are responsible for a positive skewness
in the distribution of the streamwise velocity increments, pointing at a close link
between intermittency and anisotropy (Warhaft 2002; Singh et al. 2012, 2014). This
phenomenological interpretation fits well with the wall structure organization at high
Reynolds number inferred by Morrison (2007) and Guala et al. (2011), suggesting
that large outer-scale motions and near-wall inner scales approaching the dissipative
range strongly interact, favouring rare dissipative events at the wall (see also Jimenez
2000; Mazellier & Vassilicos 2008). From a historical perspective (for details, see
Frisch 1995), strong scale interactions are inferred to be driven by inhomogeneity
and anisotropy at the scale of the turbulent kinetic energy production, which, in
this specific context, are identified as the large-scale structures of the turbulent
boundary layer. Such inhomogeneities at the large scales represent the main criticism
to the Kolmogorov (1941) universal assumptions (see e.g. Kraichnan 1974), and
the phenomenological interpretation of the observed near-wall intermittency. Despite
decades of ongoing research, the connection between the structure of wall turbulence,
near-wall anisotropy and intermittency that leads to nonlinear interactions between
large- and small-scale flow structures remains poorly understood. We hope to shed
some light with this contribution, originating from an unanticipated experimental
observation in Singh et al. (2014). In such previous work, devoted to the study
of the structure of the flow in the wake of a model wind turbine, we evidenced
an anomaly, in the form of a hump, in the asymmetry of the distribution of the,
scale-dependent, increments of the streamwise velocity. Surprisingly, the anomaly
appeared not in the turbine wake flow, but rather in the baseline flow, i.e. the
zero-pressure-gradient turbulent boundary layer achieved in the St. Anthony Falls
Laboratory (SAFL) atmospheric wind tunnel. The observed hump was located in
the frequency and amplitude ranges where Basu et al. (2007) observed, and barely
acknowledged, a similar feature (using the Kunkel & Marusic (2006) data in the
atmospheric surface layer at Surface Layer Turbulence and Environmental Science
Asymmetry in turbulent boundary layer flows 551
Test (SLTEST) facility). Singh et al. (2014) interpreted the anomaly as a peak in the
scale coupling activity, occurring in a range of scales between the Kolmogorov scale
and the integral time scale (ITS), in fact covering nearly the entire inertial range.
In this work, in addition to SAFL wind tunnel measurements, we use the SLTEST
measurements published in Metzger, McKeon & Holmes (2007) and Guala et al.
(2011) to confirm the occurrence, and investigate the scaling, of the asymmetry hump,
aiming to provide a more in-depth interpretation of such anomaly, inferred here to
be a universal, genuine, feature of turbulent boundary layer flows. The added benefit
of this specific high-Reynolds-number dataset lies in the multi-hot-wire simultaneous
acquisition providing (i) high vertical resolution (via 29 sensors logarithmically
spaced up to 5 m above the ground level), (ii) statistical convergence of the mean
and the root-mean-square (r.m.s.) streamwise velocities (Guala et al. 2011), and
(iii) strictly neutral thermal stability conditions (z/L < 0.01 (Metzger et al. 2007),
where z indicates the measurement height and L the Monin Obhukov length). The
comparative analysis across the two datasets ensures that the wall-normal scaling
arguments are independent of the Reynolds number, which strengthens the importance
of the SLTEST data to avoid scaling ambiguities: high Reynolds number uτ δ/ν
implies that the outer scale (δ) and the inner, viscous scale (ν/uτ ) are well separated
(here uτ is the shear velocity, δ is the boundary layer height and ν is the kinematic
viscosity). The asymmetry index provides the additional benefit to be a compact,
scale-dependent, height-dependent variable that can be used also as a probe to infer
how the structural population of turbulent boundary layer flows evolve with the
Reynolds number.
The paper is structured as follows. In the following section we briefly outline the
two experimental set-ups and the data used in this study. Section 3 presents major
results based on magnitude cumulant analysis and asymmetry index, providing inner
and outer scaling arguments for the two different Reynolds-number flows studied here.
Finally, the summary and conclusions drawn are presented in § 4.
2. Brief description of the datasets analysed

We present here two datasets, from wind tunnel (WT) and atmospheric surface layer
SLTEST (SLT), each comprising a profile of instantaneous streamwise velocity time
series sampled at high resolution with hot-wire probes. Both datasets were obtained in
nearly smooth zero-pressure-gradient turbulent boundary layer flows at very different
Reynolds numbers, Reτ = uτ δ/ν ' 6.7 × 105 (SLT) and 1.3 × 104 (WT). The SLT data
were captured at the SLTEST site in the Great Salt Lake Flats in Utah, a renowned
field for near-wall atmospheric boundary layer studies (Metzger & Klewicki 2001;
Smits et al. 2011, and references therein). The current dataset was acquired in May
2005 and extensively presented and discussed in Metzger et al. (2007) and Guala
et al. (2010, 2011). A total of 29 logarithmically spaced hot-wire probes were
used to sample at 5 kHz the streamwise velocity up to a height of ≈5 m in the
atmospheric surface layer. The particularly favourable conditions (neat, monitored
transition between the convective and the stable stratified boundary layer with
|z/L| < 0.01, persistence in the wind magnitude and direction) allowed this dataset
to be representative of the canonical high-Reynolds-number turbulent boundary layer
flow. The WT data were obtained in the SAFL atmospheric, temperature-controlled,
wind tunnel (Carper & Porté-Agel 2008; Chamorro & Porté-Agel 2009; Singh et al.
2014; Howard et al. 2015a,b). The main test section in the wind tunnel is 16 m long,
1.7 m wide and 1.7 m high, with a picket fence trip allowing the boundary layer
(a) 10
8
4
0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
(b) 5
4
3
2
1
0 1 2 3 4 5 6 7 8 9 10
F IGURE 1. Time series of streamwise flow velocities sampled at a frequency of 10 000 Hz

for the case of wind tunnel WT (a) and at a frequency of 5000 Hz for the case of
SLTEST SLT (b). Measurements were acquired at comparable heights in wall units, z+ =
1587 (WT) and z+ = 1464 (SLT).
to grow up to approximately δ ' 0.6 m in the available 16 m fetch. The floor and
air temperatures were controlled separately, allowing the implementation of desired
thermal stratification conditions in the boundary layer or a constant temperature
profile (neutral conditions), as in this experiment. The WT data were acquired for
baseline flow characterization in the Singh et al. (2014) study on wind turbine model
wakes. A cross-wire anemometer was used to measure instantaneous streamwise and
vertical flow velocities at a frequency of 10 kHz, for a total acquisition time of 300 s
(3 × 106 points) per vertical location. Measurement points were spaced 0.01 m in the
vertical direction from 0.004 m to 0.7 m.
Figure 1 shows the typical time series of streamwise velocities for the WT
(figure 1a) and for the SLT (figure 1b) collected at a normalized vertical distance
of z+ ∼ 1500 (z+ = zuτ /ν). Figure 2(a) shows the mean velocity profile for the WT
(circles) and the SLT (diamonds) datasets, while table 1 lists the main characteristics
of the two boundary layer flows. The streamwise r.m.s. velocities, normalized by uτ
for both WT and SLT, along with Reynolds stress for WT can be seen in figure 2(b).
As exhaustively discussed in Metzger et al. (2007) and Guala et al. (2011), we
acknowledge that the SLT mean velocity profile manifests a weak contamination
from roughness effects, with a sand-equivalent roughness estimated as ks+ ≈ 25; this
implies that results obtained at z+ < 3ks+ ∼ 75 must be taken with some caution.
Also note that the apparent WT–SLT discrepancy in the streamwise velocity turbulent
intensity profile urms (z+ ) for z+ > 3 × 102 is a clear effect of the increased Reynolds
number (Metzger et al. 2007).
3. Results and discussion

3.1. Magnitude cumulant analysis: a measure of intermittency
Spatial or temporal series of velocity fluctuations in turbulent flows are typically
decomposed into Fourier modes to compute the power spectral density that describes
the energetic contribution of eddies of different sizes, or, in statistical terms, identifies
scale-dependent variations of the second-order moment, or variance (Singh, Porté-Agel
& Foufoula-Georgiou 2010; Singh et al. 2014). In addition, the power spectrum
(a) 35 (b) 3.5

30 3.0
25 2.5
20 2.0
15 1.5
WT
10 SLT 1.0
Osterlund
5 Eaton 0.5
0 0
101 102 103 104 105 101 102 103 104 105
F IGURE 2. (Colour online) Mean (a) and fluctuating (b) streamwise velocity statistics
for the WT and SLT datasets, normalized with the shear velocity uτ . Inclined lines in
(a) indicate the logarithmic mean velocity profile for the smooth case and for increasing
roughness ks+ = 5, 10, 25, 35 (for reference). The + symbols refer to measurements from
De Graaf & Eaton (2000) at Reτ = 1430 and Osterlund (1999) at Reτ = 2530. Subscript
rms in (b) represents root-mean-square velocity.
Dataset U∞ uτ δ Reτ Ttotal Np z+ z/δ

(m s−1 ) (m s−1 ) (m) (s)
WT 7.12 0.35 0.6 1.3 × 104 300 35 70–9000 0.02–1
SLT ≈6 0.15 ≈80 6.7 × 105 240 29 10–46 000 0.002–0.06
TABLE 1. Mean flow statistics for the wind tunnel (WT) and atmospheric surface layer
(SLT) datasets: Ttotal indicates the total sampling time at each measurement location, while
Np is the number of probe locations in the vertical direction. The ranges of heights covered
in the two datasets are reported in viscous (z+ ) and outer units (z/δ).
provides all the necessary information for re-creating a representative turbulent

velocity time series in the case when the velocity increments are Gaussian-distributed
over a range of scales (Singh et al. 2011, 2012); the velocity increments are defined
here as 1u(t, a) = u(t + a) − u(t) as a function of the separation time scale a.
However, the distribution of the velocity increments in a turbulent signal usually
shows significant deviation from the Gaussian distribution at smaller scales, i.e. for
values of a approaching the Kolmogorov time scale τk (see e.g. Malecot et al. 2000;
Ruiz-Chavarria et al. 2000). This can also be observed in the probability distribution
function (p.d.f.) of velocity increments for WT (figure 3a) and SLT (figure 3b),
suggesting (i) a multiscaling type of behaviour of turbulent velocity fluctuations,
and (ii) that higher order statistical moments are required for reconstructing the
original velocity signal. Note that the Kolmogorov time scale is defined here as
τk = η/U, where η = (ν 3 /)1/4 is the Kolmogorov length scale, is the average rate
of dissipation of turbulent kinetic energy, ν is the kinematic viscosity and U is the
average convective velocity. This is an advective form of the Kolmogorov time scale,
which is different from the Kolmogorov time scale τ = (ν/)1/2 used in homogeneous
turbulence and inadequate when the convective velocity is much larger than the
(a) 100 (b) 100
10–1 10–1
P.d.f.
10–2 10–2
10–3 10–3
–5 0 5 –5 0 5
F IGURE 3. (Colour online) Semilogarithmic p.d.f.s of the normalized form of the velocity
increments 1u(t, a) = u(t + a) − u(t)) for the scales a = τk and a = ITS for both WT (a)
and SLT (b) datasets. The solid curve denotes the Gaussian distribution. P.d.f.s are shown
for z+ +
WT = 1587 (a) and zSLT = 1464 (b).
Kolmogorov velocity. The ITS is estimated using the integral of the autocorrelation
function of the streamwise velocity.
While a large amount of work has been devoted to high-order structure function
analysis (Menevau & Sreenivasan 1991; Toschi, Leveque & Ruiz-Chavarria 2000;
Venugopal et al. 2006; Basu et al. 2007; Kholmyansky, Moriconi & Tsinober 2007;
Singh et al. 2009, 2011; Guala et al. 2010), an alternative method has been proposed
in the form of the magnitude cumulant analysis, introduced by Delour, Muzy &
Arnéodo (2001) – see also Malecot et al. (2000) and Chevillard et al. (2005). This
approach requires only the second-order magnitude cumulant to unravel the underlying
intermittency in the turbulence structure of the velocity signal, under the assumption
of log-normality in the increment distribution (see e.g. Delour et al. 2001; Chevillard
et al. 2005; Basu et al. 2007; Singh et al. 2014). In general, for a random variable X
having a p.d.f. F(x), the moment generating function (also defined as the characteristic
function) can be expressed as
Z +∞
φp (k) = eikx F(x) dx, (3.1)
−∞
which can be expanded in a Taylor’s series as the following and leading to

∞
(ikx)2 (ikx)3 (ik)n
X
ikx
φp (k) = he i = 1 + ikx + + + ··· = Mn , (3.2)
2! 3! n=0
n!
where Mn is the nth-order moment of X, represented as

Z +∞
Mn = xn F(x) dx, n = 0, 1, 2, . . . . (3.3)
−∞
Analogously, Mn can be estimated by the nth-order derivative of φp (k) at k = 0. The

cumulant generating function of X can be accordingly defined as
ψp (k) ≡ ln φp (k) = lnheikx i, (3.4)

while the cumulant Cn of X can be obtained from the nth derivative of ψp (k).
Moments and cumulants can be related as
C1 = M1 , (3.5)
C2 = M2 − M12 , (3.6)
C3 = M3 − 3M2 M1 + 2M13 , (3.7)
C4 = M4 − 4M3 M1 − 3M22 + 12M2 M12 − 6M14 , (3.8)
and so on.
If we substitute the variable X with the velocity increment 1u separated by a given
lag a, the magnitude cumulant Cn can be expressed as
∞
!
q
X qn
h|1a u| i = exp Cn (a) , (3.9)
n=1
n!
where
C1 = hln |1u|i ∼ c1 ln(a), (3.10)

C2 = h(ln |1u|)2 i − hln |1u|i2 ∼ −c2 ln(a), (3.11)
C3 = h(ln |1u|)3 i − 3h(ln |1u|)2 ihln |1u|i + 2hln |1u|i3 ∼ c3 ln(a). (3.12)
Cumulants C1 and C2 are scale-dependent functions of the mean hln |1u|i and
the variance h(ln |1u|)2 i (for more details, see Malecot et al. 2000; Delour et al.
2001). Figure 4(a) shows the second-order cumulant C2 as a function of scale a
normalized by the ITS for both WT and SLT (see also Singh et al. (2014) for the
plot of C1 (a), whose slope from log-linear regime represents the Hölder exponent hhi,
for the WT dataset). The slope of the log-linear regime obtained from second-order
cumulant C2 (a) within the range of scales of interest (from Kolmogorov scale to
integral time scale, τk < a < ITS), is used to define the intermittency coefficient
c2 ; c2 represents a compact, scale-independent, parameter describing the structure
of each streamwise velocity increment time series, by providing a measure of the
inhomogeneous temporal arrangement of its local abrupt fluctuations (Singh et al.
2011, 2012) and is not to be confused with the above-defined second-order cumulant
C2 notation.
Figure 4(b) shows the computed intermittency coefficient c2 profiles for the WT and
SLT cases, for the range of wall-normal distances z+ explored here. It is noted that the
estimated intermittency coefficients c2 obtained from traditional high-order structure
function analysis (not shown here for brevity) resulted in similar values as obtained
from the magnitude cumulant analysis presented here. As can be seen from figure 4(b),
the c2 profiles for both WT and SLT are very similar and, in fact, overlap for a range
of z+ , particularly in the range z+ ∼ 70–103 . For the lower z+ , SLTEST data show
an increase in c2 with decreasing z+ , suggesting an increase in the frequency and
magnitude of the intense velocity gradients that may have resulted from increasing
mean shear as z+ decreases (see also Antonia, Orlandi & Romano 1998; Onorato,
Camussi & Iuso 2000). Note that, in the case of a monofractal signal, where the
shape of the increment distributions of the signal does not change with the scale a, the
intermittency coefficient c2 is precisely 0. Also, note that the c2 values in our study for
both WT and SLT datasets are consistent with the c2 observed in Basu et al. (2007)
for a similar dataset in the atmospheric surface layer.
(a) 1.55 (b) 10 5

WT
1.50 SLT
1.45 10 4
1.40
1.35 10 3
1.30
1.25 10 2
1.20
1.15 101
10–4 10–2 100 102 0 0.05 0.10 0.15 0.20
F IGURE 4. (a) Second-order cumulant C2 as a function of scale a normalized by the

ITS and (b) vertical profile of intermittency coefficients (c2 ) as a function of vertical wall
distance z+ for the streamwise velocity for both WT (stars) and SLT (circles). Note that
the intermittency coefficients c2 shown in figure 4(b) are computed from the second-order
cumulants C2 (figure 4a) for a range of scales, i.e. Kolmogorov scale τk to ITS.
WT
4 SLT
–2
–15 –10 –5 0 5 10 15
x
F IGURE 5. Fifth moment functions (x5 f (x), where x = 1u(t, a)/h(1u(t, a))2 i1/2 ) of the
streamwise velocity increments for the WT and the SLT. Note that the p.d.f.s of the
velocity increments for both WT and SLT were computed for a = τk , at z+ WT = 1587 and
z+
SLT = 1464, respectively.
3.2. Scale-dependent asymmetry of the velocity increments

From figure 3 it was noticed that the p.d.f.s of the velocity increments, for varying
time scale a, are neither Gaussian nor symmetric, a combination that is inferred to
point at multiplicative cascade processes and non-local scale interactions (Schertzer
et al. 1997; Warhaft 2002; Singh et al. 2012, 2014). While non-Gaussianity is
expressed through the intermittency coefficient c2 , the asymmetry in the distribution
can be quantified using either the third or the fifth moment function (as in this study)
of the velocity increments (see Warhaft 2002). Results are provided in figure 5 for
WT and SLT velocity time series at z+ ∼ 1500 for the smallest scale separation
a = τk . The qualitative behaviour of the two curves directs attention towards a
Reynolds-number-independent feature of the turbulent boundary layer, regarding the
distribution and statistical weight of rare but intense dissipative events.
(a) 783 453 (b) 28 615

228 3216 680 9230 34 389
0.30 505 907
0.30 16 567
634 227 3173 23 846
0.25 0.25
0.20 0.20
0.15 0.15
0.10 0.10
0.05 0.05
0 0
–0.05 –0.05
10–4 10–2 100 102 10–4 10–2 100 102
F IGURE 6. (Colour online) Asymmetry index of the streamwise velocity increments as a

function of scale for WT (black lines) and SLT (blue lines) at 70 < z+ < 3.3 × 103 (a)
and at z+ > 4 × 104 (b). The x-axis represents the time scale a normalized by the ITS.
To better quantify the scale-dependent asymmetry of the velocity increment

distribution, we define the asymmetry index as a function of the time scale a and
vertical height z, namely A(a, z) = h(X − X)3 i/h|(X − X)3 |i, where X = 1u(a, z) and X
is the mean of X. The asymmetry index is plotted in figure 6(a) for different scales,
at different heights, for both WT and SLT datasets. As outlined in the introduction, a
hump is observed in a range of scales smaller than the ITS in both datasets, consistent
in terms of frequency and amplitude with the hump reported by Basu et al. (2007).
Here we confirm the location of the hump in the inertial range using an independent
dataset from atmospheric surface layer (SLT) and a new wind tunnel experiment
(WT).
3.3. Reynolds-number-independent asymmetry in turbulent boundary layers

As the difference in Reynolds number between SLTEST and the SAFL wind tunnel
is of the order of two decades, the emerging hump-like feature might be universal
to turbulent boundary layer flows. One of the main objectives of this paper is to
compare systematically the amplitude and location of the hump at varying wall-normal
distances, providing different normalization options for both the vertical coordinate z
and the velocity increment time scale a. The main hypothesis put forward is how the
asymmetry index evolves in different regions of the turbulent boundary layer, for the
two available Reynolds-number experiments. In order to allow a direct comparison
between the laboratory-scale (WT) and field-scale (SLT) datasets, we kept a consistent
normalization for a, using the integral time scale for normalization, a/ITS.
As discussed earlier in the introduction, scale interactions near the wall are the
main source of intermittency and anisotropy observed in high-order statistics of the
streamwise velocity (Guala et al. 2010, 2011; Singh et al. 2014). In the continuous
cycle of mutual influence between (inner) near-wall turbulence and (outer) δ-scale
motions, we argue that a single scaling may not work across the whole range
of turbulent scales and wall distances. For this purpose, we investigate here if
inner (z+ = zuτ /ν) or outer (z/δ) unit normalization will lead to a collapse of the
0.40
0.35 20
28
0.30 44
76
0.25 228
0.20
0.15
0.10
0.05
–0.05
10–3 10–2 10–1 100 101 102
F IGURE 7. (Colour online) Asymmetry index of the streamwise velocity increments as

a function of scale for WT (black line) and SLT (blue lines) for z+ < 102 . The x-axis
represents the time scale a normalized by the ITS.
anisotropy (asymmetry index) curves A(a, z) for a specific range of temporal scales a
and vertical locations z. In this respect, we recall that the spacing between near-wall
streaks ('100 viscous wall units) has been proved to be fairly invariant across a wide
range of Reynolds number (Metzger & Klewicki 2001). Recognizing the key role of
near-wall structures to inner–outer scale interaction processes, we first attempt to use
the viscous scale ν/u∗ for wall distance normalization. We compare scale-dependent
asymmetry curves for the two datasets at the fixed z+ location in figure 6(a,b). Here
the normalizing time scale for the velocity increment is kept as ITS, which of course
varies in the two Reynolds-number flows.
In figure 6(a), we note how the hump maintains its amplitude and temporal
location for a wide range of z+ , independent of the Reynolds number, unambiguously
marking a statistical property of turbulent boundary layer flows. However, for
z+ > 4000 (figure 6b), we start to notice appreciable deviations for large time
scales a approaching the ITS, suggesting a change in scale interaction processes, and
possibly in wall turbulence organization at the SLT high-Reynolds-number flow. For
even larger wall distances, inspected in figure 6(b), the asymmetry peak is lost within
ample oscillations that might be the result of the interactions between large- and
very-large-scale motions populating the logarithmic layer of the SLT flow; in such a
case, we must note that the SLT dataset may not be long enough to ensure statistical
convergence of the scale-dependent asymmetry index in the full temporal–vertical
domain investigated. This said, figure 6(b) confirms that the flow region where
asymmetry statistics overlap is confined in the O(102 –103 ) range of viscous wall
units. Note that an independent set of data from a wind tunnel experiment (not
shown here) showed similar trends and results to those presented here.
The SLT data provide the opportunity to further explore the scale-dependent
asymmetry very close to the wall (z+ < 100; see figure 7). As the surface is
approached, the amplification of A(a, z) in the Kolmogorov scale range contributes
to increase the peak value while smearing the hump signature. With intermittency
and small-scale asymmetry both high at the wall, scale coupling is inferred to occur
within a different layer and population of wall turbulent structures, as compared to
the case discussed above in the logarithmic layer. Following a structural interpretation,
figure 7 suggests a strong effect of large- and very-large-scale motions at the surface,
where the mean shear is strongest, literally sweeping the wall and distorting the
distribution of the near-wall streamwise velocity at those small scales that were
traditionally marked as homogeneous and isotropic. Recently Hultmark et al. (2013)
quantified third-order central moments of velocity fluctuations obtained at high
Reynolds number and observed increasing skewness with decreasing wall-normal
distance. These observations are in qualitative agreement with our results, suggesting
that scale coupling increases as the wall distance decreases. The evolution of the
asymmetry peak appears to occur quite sharply in the (a, z) domain, suggesting a
marked change in the wall structure organization. This is consistent with the spatial
transition between the near-wall streaks and the ramp-like structure organization above
z+ ∼ 70, which marks the lower end of the logarithmic layer. With increasing distance
from the wall (10 < z+ < 50), the streaks’ asymmetry is reduced at the small time
scales, until the hump z+ > 70 clearly emerges in the temporal (inertial range) and
wall-normal domain of the attached eddies (see Perry, Henbest & Chong 1986).
The A(a, z) vertical distribution suggests that local high shear induced by large-scale
structures impinging the wall is felt by near-wall streaks, leading to increased
intermittency (see Guala et al. 2010, and also figure 4b), asymmetry and non-local
energy transfer (Warhaft 2002; Singh et al. 2014). The asymmetry profiles also
indicate that the ramp-like structures inferred in Guala et al. (2011) to populate, at
the atmospheric scale (high Reynolds number), a relatively near-wall region of the
boundary layer up to z+ ∼ 5 × 103 display the asymmetry-hump signature in the
same vertical range. The role of ramp-like structures is confirmed by employing the
attached eddy time scale z/u∗ as a replacement for the ITS: in figure 8(a), asymmetry
curves in the near-wall region are presented as a function of a/(z/u∗ ) and observed to
partially overlap across a range of scales consistent with those reported in figure 6(a).
The validity of inner, viscous, scaling and its conceptual implications are further
evidenced in figure 8(b). When asymmetry index curves are compared for both WT
and SLT at different z+ but similar z/δ locations, no significant data collapse is
observed. Near-wall scaling is thus confirmed in a region well above the near-wall
spectral peak of Hutchins & Marusic (2007a) and well within the logarithmic layer,
where outer scaling would have been, in principle, an equally valid option. While we
acknowledge that a full comparison should be carried out using measurements in the
outer layer at the SLTEST Reynolds number (not available here), we must recognize
that strong scale interactions, intermittency and anisotropy are typically observed
in high-shear flows, thus naturally residing in the near-wall region. Therefore, it is
speculated that outer scale normalization for A(a, z) will hardly apply in large-scale
geophysical flows, and that the asymmetry hump will remain a near-wall turbulence
signature, with near wall defined in viscous units.
4. Summary and conclusions

The distribution of scale-dependent streamwise velocity increment 1u(a, z) =
u(t + a, z) − u(t, z) was investigated at varying wall-normal locations in turbulent
boundary layer flows at laboratory and atmospheric Reynolds numbers, using
measurements from the SAFL wind tunnel and the SLTEST site (Metzger et al.
(a) 783 453 (b) 0.063 0.05

228 3216 680 0.014 0.067
505 907 0.030 0.017
634 227 3173 0.052 0.033
0.30 0.30
0.25 0.25
0.20 0.20
0.15 0.15
0.10 0.10
0.05 0.05
0 0
–0.05 –0.05
10–4 10–2 100 102 10–4 10–2 100 102
F IGURE 8. (Colour online) (a) Asymmetry index of the streamwise velocity increments as
a function of scale for WT (black lines) and SLT (blue lines) at z+ < 103 . (b) Asymmetry
index of the streamwise velocity increments as a function of scale for WT (black lines)
and SLT (blue lines) for comparable z/δ locations. The x-axis in (a) represents the time
scale a normalized by z/u∗ , whereas in (b) it represents the time scale a normalized by
the ITS.
2007). The fifth-order moment of the velocity increments was employed here to
emphasize the scale-dependent non-Gaussianity of the increment distribution, leading
to the definition of the scale-dependent asymmetry index
h(1u(a, z) − 1u(a, z))3 i

A(a, z) = . (4.1)
h|(1u(a, z) − 1u(a, z))3 |i
The shape of A(a, z) revealed a well-defined hump located in the inertial range,
thus for time scales a smaller than the ITS and larger than the Kolmogorov scale
τk . The hump is observed for z+ < 5 × 103 , with peak location at time scale
a/ITS ∼ a/(z/u∗ ) ∼ 0.1, and amplitude A(a, z) ∼ 0.2, with a tendency to shift towards
smaller time scales as the surface is approached (z+ < 102 ). The asymmetry hump
location in the temporal and vertical domain is observed to obey inner wall scaling,
while outer-scale normalization, i.e. comparing A(a, z) curves at similar z/δ, did not
seem to provide any useful insight. By comparing the two datasets, the hump in
A(a, z) emerged as a robust, statistical feature of the canonical turbulent boundary
flow, independent of the Reynolds number.
A parallel analysis using magnitude cumulant analysis of the scale-dependent
velocity increments revealed that intermittency is also enhanced near the wall, well
within the logarithmic region, in the same domain where the hump signature was
observed. The combination of asymmetry and intermittency is inferred to point at
non-local energy transfer, i.e. between non-neighbouring scales in the frequency
domain. From a turbulent structure perspective this is an indication of strong scale
interaction processes (Warhaft 2002; Guala et al. 2011): outer-scale motions in
the logarithmic layer are inferred to impact and distort small scales near the wall,
favouring abrupt energy transfer across scales, bypassing the typically assumed energy
cascade of the inertial range, and causing rare but highly dissipative events that leave
a mark on high-order statistics. The asymmetry peak at small scales for z+ < 70
suggests that near-wall streaks are mostly affected; in addition, evidence of the
asymmetry hump up to z+ ∼ 5000 implies that also ramp-like structures are likely
to be modulated, denoting a wide range of multi-scale interactions in a significant
portion of the wall region. The fact that the attached-eddy scaling z/u∗ and inner
normalization (z+ ) lead to a collapse of the A(a, z) curves further suggests that
ramp-like structures probably obey near-wall scaling as opposed to outer scaling, as
inferred in Guala et al. (2011).
We acknowledge that the observed oscillations of the asymmetry index for large
time scales, estimated from the atmospheric (SLT) dataset, may be due to the finite
length of the velocity time series that may affect the convergence of high-order
statistics at large scales. In addition, the increased intermittency observed near the
wall in the SLT data, and (to some extent) far from the wall in the WT data,
points at a broader question on the effect of the intensity and vertical distribution
of the mean shear on the structure of turbulence. Intermittency has been observed
to be enhanced in shear flows as compared to homogeneous isotropic turbulence,
exhibiting some (Antonia et al. 1998; Onorato et al. 2000) or minimal (Toschi et al.
2000) wall-normal dependence. Ruiz-Chavarria et al. (2000) evidenced that different
scaling regions in the structure functions can affect the estimates of height-dependent
and scale-dependent intermittency. Additionally, measurements in the atmospheric
boundary layer at O(102 ) m over urban or coastline areas revealed strong intermittency
at large (time) scales (Boettcher et al. 2003; Liu et al. 2010).
For all these reasons, it would be important to confirm (or discuss) the trends
outlined in this paper using other datasets from canonical boundary layer flows at
high Reynolds number.
Acknowledgements
We thank Professors M. Metzger and B. J. McKeon for providing the SLTEST
data. We also thank the editor, I. Marusic, and the three reviewers, whose suggestions
and constructive comments substantially improved our presentation and refined our
interpretations. This research has been supported by our own curiosity and partially by
National Science Foundation (NSF) Career Grant Geophysical Flow Control to M.G.
REFERENCES
A NTONIA , R. A., O RLANDI , P. & ROMANO , G. P. 1998 Scaling of longitudinal velocity increments
in a fully developed turbulent channel flow. Phys. Fluids 10, 3239.
B ASU , S., F OUFOULA -G EORGIOU , E., L ASHERMES , B. & A RNÉODO , A. 2007 Estimating
intermittency exponent in neutrally stratified atmospheric surface layer flows: a robust
framework based on magnitude cumulant and surrogate analyses. Phys. Fluids 19, 115102.
B ENZI , R., C ILIBERTO , S., T RIPICCIONE , C., B AUDET, C., M ASSAIOLI , F. & S UCCI , F. 1993
Extended self similarity in turbulent flows. Phys Rev. E 48, R29–R32.
B OETTCHER , F., R ENNER , C. H., WALDL , H. P. & P EINKE , J. 2003 On the statistics of wind gusts.
Boundary-Layer Meteorol. 108, 163–173.
C ARPER , M. & P ORTÉ -AGEL , F. 2008 Subfilter-scale fluxes over a surface roughness transition. Part
I: measured fluxes and energy transfer rates. Boundary-Layer Meteorol. 126, 157–179.
C HAMORRO , L. P. & P ORTÉ -AGEL , F. 2009 A wind-tunnel investigation of wind-turbine wakes:
boundary-layer turbulence effects. Boundary-Layer Meteorol. 132, 129–149.
C HEVILLARD , L., ROUX , S. G., L EVEQUE , E., M ORDANT, N., P INTON , J.-F. & A RNEODO , A.
2005 Intermittency of velocity time increments in turbulence. Phys. Rev. Lett. 95, 064501.
D E G RAAFF , D. B. & E ATON , J. K. 2000 Reynolds-number scaling of the flat-plate turbulent
boundary layer. J. Fluid Mech. 422, 319–346.
D ELOUR , J., M UZY, J. F. & A RNÉODO , A. 2001 Intermittency of 1D velocity spatial profiles in
turbulence: a magnitude cumulant analysis. Eur. Phys. J. B 23, 243–248.
F RISCH , U. 1995 Turbulence, the Legacy of A. N. Kolmogorov. Cambridge University Press.
G UALA , M., H OMMEMA , S. E. & A DRIAN , R. J. 2006 Large-scale and very-large-scale motions in
turbulent pipe flow. J. Fluid Mech. 554, 521–542.
G UALA , M., M ETZGER , M. & M C K EON , B. J. 2010 Intermittency in the atmospheric surface layer:
unresolved or slowly varying? Physica D 239 (14), 1251–1257.
G UALA , M., M ETZGER , M. & M C K EON , B. J. 2011 Interactions within the turbulent boundary layer
at high Reynolds number. J. Fluid Mech. 666, 573–604.
J IMENEZ , J. 2000 Intermittency and cascades. J. Fluid Mech. 409, 99–120.
H OWARD , K. B., H U , J. S., C HAMORRO , L. P. & G UALA , M. 2015a Characterizing the response
of a wind-turbine model under complex inflow conditions. Wind Energy 18 (4), 729–743.
H OWARD , K. B., S INGH , A., S OTIROPOULOS , F. & G UALA , M. 2015b On the statistics of wind
turbine wake meandering: an experimental investigation. Phys. Fluids 27 (7), 075103.
H ULTMARK , M., VALLIKIVI , M., BAILEY, S. C. C. & S MITS , A. J. 2013 Logarithmic scaling of
turbulence in smooth- and rough-wall pipe flow. J. Fluid Mech. 728, 376–395.
H UTCHINS , N. & M ARUSIC , I. 2007a Large-scale influences in near-wall turbulence. Phil. Trans. R.
Soc. Lond. A 365, 647–664.
H UTCHINS , N. & M ARUSIC , I. 2007b Evidence of very long meandering features in the logarithmic
region of turbulent boundary layers. J. Fluid Mech. 579, 1–28.
K EYLOCK , C. J., S INGH , A. & F OUFOULA -G EORGIOU , E. 2013 The influence of migrating bed
forms on the velocity–intermittency structure of turbulent flow over a gravel bed. Geophys.
Res. Lett. 40, 1351–1355.
K EYLOCK , C. J., S INGH , A., V ENDITTI , J. G. & F OUFOULA -G EORGIOU , E. 2014 Robust
classification for the joint velocity–intermittency structure of turbulent flow over fixed and
mobile bedforms. Earth Surf. Process. Landf. 39, 1717–1728.
K OLMOGOROV, A. N. 1941 Local structure of turbulence in an incompressible liquid for very large
Reynolds numbers. C. R. Acad. Sci. USSR 30, 301–305.
K HOLMYANSKY, M., M ORICONI , L. & T SINOBER , A. 2007 Large scale intermittency in the
atmospheric boundary layer. Phys. Rev. E 76, 026307.
K IM , K. C. & A DRIAN , R. J. 1999 Very large-scale motion in the outer layer. Phys. Fluids 11,
417–422.
K RAICHNAN , R. H. 1974 On Kolmogorov inertial range theories. J. Fluid Mech. 62 (2), 306–330.
K UNKEL , G. J. & M ARUSIC , I. 2006 Study of the near-wall-turbulent region of the high-Reynolds-
number boundary layer using an atmospheric flow. J. Fluid Mech. 548, 375–402.
L AVAL , J.-P., D UBRULLE , B. & N AZARENKO , S. 2001 Nonlocality and intermittency in
three-dimensional turbulence. Phys. Fluids 13, 1995.
L IU , L., H U , F., C HENG , X.-L. & S ONG , L.-L. 2010 Probability density functions of velocity
increments in the atmospheric boundary layer. Boundary-Layer Meteorol. 134, 243–255.
M ALECOT, Y., AURIAULT, C., K AHALERRAS , H., G AGNE , Y., C HANAL , O., C HABAUD , B. &
C ASTAING , B. 2000 A statistical estimator of turbulence intermittency in physical and numerical
experiments. Eur. Phys. J. B 16, 549–561.
M ARUSIC , I., M ATHIS , R. & H UTCHINS , N. 2010 Predictive model for wall-bounded turbulent flow.
Science 329 (5988), 193–196.
M ATHIS , R., H UTCHINS , N. & M ARUSIC , I. 2009 Large-scale amplitude modulation of the small-scale
structures in turbulent boundary layers. J. Fluid Mech. 628, 311–337.
M AZELLIER , N. & VASSILICOS , J. C. 2008 The turbulence dissipation constant is not universal
because of its universal dependence on large-scale flow topology. Phys. Fluids 20, 015101.
M ENEVAU , C. & S REENIVASAN , K. R. 1991 The multifractal nature of turbulent energy dissipation.
J. Fluid Mech. 224, 429–484.
M ETZGER , M. M. & K LEWICKI , J. C. 2001 A comparative study of near-wall turbulence in high
and low Reynolds number boundary layers. Phys. Fluids 13, 692–701.
M ETZGER , M., M C K EON , B. J. & H OLMES , H. 2007 The near-neutral atmospheric surface layer:
turbulence and non-stationarity. Phil. Trans. R. Soc. Lond. A 365, 859–876.
M ININNI , P. D., A LEXAKIS , A. & P OUQUET, A. 2006 Large-scale flow effects, energy transfer, and
self-similarity on turbulence. Phys. Rev. E 74, 016303.
M ORRISON , J. F. 2007 The interaction between inner and outer regions of turbulent wall-bounded
flows. Phil. Trans. R. Soc. Lond. A 365, 683–698.
O NORATO , M., C AMUSSI , R. & I USO , G. 2000 Small scale intermittency and bursting in a turbulent
channel flow. Phys. Rev. E 61 (2), 1446–1454.
O STERLUND , J. M. 1999 Experimental studies of zero pressure-gradient turbulent boundary-layer
flow. PhD thesis, Department of Mechanics, Royal Institute of Technology, Stockholm.
P ERRY, A. E., H ENBEST, S. & C HONG , M. S. 1986 A theoretical and experimental study of wall
turbulence. J. Fluid Mech. 165, 163–199.
P OGGI , D., P ORPORATO , A., R IDOLFI , L., A LBERTSON , J. D. & K ATUL , G. 2004 Interaction between
large and small scales in the canopy sublayer. Geophys. Res. Lett. 31 (5), L05102.
RUIZ -C HAVARRIA , G., C ILIBERTO , S., BAUDET, C. & L ÉVÊQUEB , E. 2000 Scaling properties of the
streamwise component of velocity in a turbulent boundary layer. Physica D 141, 183–198.
S CHERTZER , D., L OVEJOY, S., S CHMITT, F., C HIGUIRINSKAYA , Y. & M ARSAN , D. 1997 Multifractal
cascade dynamics and turbulent intermittency. Fractals 5, 427–471.
DE S ILVA , C. M., M ARUSIC , I., W OODCOCK , J. D. & M ENEVEAU , C. 2015 Scaling of second- and
higher-order structure functions in turbulent boundary layers. J. Fluid Mech. 769, 654–686.
S INGH , A., F IENBERG , K., J EROLMACK , D. J., M ARR , J. & F OUFOULA -G EORGIOU , E. 2009
Experimental evidence for statistical scaling and intermittency in sediment transport rates.
J. Geophys. Res. 114, F01025.
S INGH , A., F OUFOULA -G EORGIOU , E., P ORTÉ -AGEL , F. & W ILCOCK , P. R. 2012 Coupled
dynamics of the co-evolution of bed topography, flow turbulence and sediment transport in
an experimental flume. J. Geophys. Res. 117, F04016.
S INGH , A., H OWARD , K. B. & G UALA , M. 2014 On the homogenization of turbulent flow structures
in the wake of a model wind turbine. Phys. Fluids 26 (2), 025103.
S INGH , A., L ANZONI , S., W ILCOCK , P. R. & F OUFOULA -G EORGIOU , E. 2011 Multi-scale statistical
characterization of migrating bedforms in gravel and sand bed rivers. Water Resour. Res. 47,
W12526.
S INGH , A., P ORTÉ -AGEL , F. & F OUFOULA -G EORGIOU , E. 2010 On the influence of gravel bed
dynamics on velocity power spectra. Water Resour. Res. 46, W04509.
S MITS , A. J., M C K EON , B. J. & M ARUSIC , I. 2011 High Reynolds number wall turbulence. Annu.
Rev. Fluid Mech. 43, 353–375.
T OSCHI , F., L EVEQUE , E. & RUIZ -C HAVARRIA , G. 2000 Shear effects in nonhomogeneous turbulence.
Phys. Rev. Lett. 85 (7), 1436–1439.
V ENUGOPAL , V., ROUX , S. G., F OUFOULA -G EORGIOU , E. & A RNÉODO , A. 2006 Revisiting
multifractality of high-resolution temporal rainfall using a wavelet-based formalism. Water
Resour. Res. 42, W06D14.
WARHAFT, Z. 2002 Turbulence in nature and in the laboratory. Proc. Natl Acad. Sci. USA 99
(Suppl. 1), 2481–2486.
Y EUNG , P. K., B RASSEUR , J. G. & WANG , Q. 1995 Dynamics of direct large-small scale couplings
in coherently forced turbulence: concurrent physical- and Fourier-space views. J. Fluid Mech.
283, 43–95.

A Measure of Scale Dependent Asymmetry in Turbulent Boundary Layer Flows Scaling and Reynolds Number Similarity

Încărcat de

Informații document

Descriere originală:

Titlu original

Drepturi de autor

Formate disponibile

Partajați acest document

Partajați sau inserați document

Opțiuni de partajare

Vi se pare util acest document?

Este necorespunzător acest conținut?

Drepturi de autor:

Formate disponibile

A Measure of Scale Dependent Asymmetry in Turbulent Boundary Layer Flows Scaling and Reynolds Number Similarity

Încărcat de

Drepturi de autor:

Formate disponibile

J. Fluid Mech. (2016), vol. 797, pp. 549–563.

c Cambridge University Press 2016 549

A measure of scale-dependent asymmetry in

Arvind Singh1, †, Kevin B. Howard2 and Michele Guala2

(Received 19 September 2015; revised 18 April 2016; accepted 19 April 2016;

The distribution of temporal scale-dependent streamwise velocity increments is

Key words: atmospheric surface layer, intermittency, wall turbulence

† Email address for correspondence: Arvind.Singh@ucf.edu

2. Brief description of the datasets analysed

F IGURE 1. Time series of streamwise flow velocities sampled at a frequency of 10 000 Hz

3. Results and discussion

(a) 35 (b) 3.5

Dataset U∞ uτ δ Reτ Ttotal Np z+ z/δ

provides all the necessary information for re-creating a representative turbulent

(a) 100 (b) 100

which can be expanded in a Taylor’s series as the following and leading to

where Mn is the nth-order moment of X, represented as

Analogously, Mn can be estimated by the nth-order derivative of φp (k) at k = 0. The

ψp (k) ≡ ln φp (k) = lnheikx i, (3.4)

C1 = hln |1u|i ∼ c1 ln(a), (3.10)

(a) 1.55 (b) 10 5

F IGURE 4. (a) Second-order cumulant C2 as a function of scale a normalized by the

3.2. Scale-dependent asymmetry of the velocity increments

(a) 783 453 (b) 28 615

F IGURE 6. (Colour online) Asymmetry index of the streamwise velocity increments as a

To better quantify the scale-dependent asymmetry of the velocity increment

3.3. Reynolds-number-independent asymmetry in turbulent boundary layers

F IGURE 7. (Colour online) Asymmetry index of the streamwise velocity increments as

4. Summary and conclusions

(a) 783 453 (b) 0.063 0.05

h(1u(a, z) − 1u(a, z))3 i

S-ar putea să vă placă și