J. Fluid Mech. (2016), vol. 798, pp. 371–397.

c Cambridge University Press 2016 371

doi:10.1017/jfm.2016.319

Three-dimensional instabilities in oscillatory

flow past elliptic cylinders

José P. Gallardo1, †, Helge I. Andersson2 and Bjørnar Pettersen1

1 Department of Marine Technology, Norwegian University of Science and Technology,
NO-7491 Trondheim, Norway
2 Department of Energy and Process Engineering, Norwegian University of Science and Technology,
NO-7491 Trondheim, Norway

(Received 10 February 2015; revised 16 February 2016; accepted 4 May 2016;

first published online 3 June 2016)

We investigate the early development of instabilities in the oscillatory viscous flow

past cylinders with elliptic cross-sections using three-dimensional direct numerical
simulations. This is a classical hydrodynamic problem for circular cylinders, but
other configurations have received only marginal attention. Computed results for
some different aspect ratios Λ from 1 : 1 to 1 : 3, all with the major axis of the
ellipse aligned in the main flow direction, show good qualitative agreement with
Hall’s stability theory (J. Fluid Mech., vol. 146, 1984, pp. 347–367), which predicts
a cusp-shaped curve for the onset of the primary instability. The three-dimensional
flow structures for aspect ratios larger than 2 : 3 resemble those of a circular cylinder,
whereas the elliptical cross-section with the lowest aspect ratio of 1 : 3 exhibits
oblate rather than tubular three-dimensional flow structures as well as a pair of
counter-rotating spanwise vortices which emerges near the tips of the ellipse. Contrary
to a circular cylinder, instabilities for an elliptic cylinder with sufficiently high
eccentricity emerge from four rather than two different locations in accordance with
the Hall theory.

Key words: separated flows, vortex dynamics, vortex instability

1. Introduction
1.1. Oscillatory viscous flows
A canonical problem in marine hydrodynamics is that of the oscillatory flow around
a cylindrical body. In its simplest form, which is used as a prototype to study
more complex cases, the flow is considered to be sinusoidal, unidirectional and
perpendicular to the axis of a circular cylinder. This periodic motion is characterized
by the Keulegan–Carpenter number KC = Um T/D (Keulegan & Carpenter 1958),
where Um is the velocity amplitude, T is the period of the motion and D is
the diameter of the circular cylinder. Viscous effects are taken into account by
a second governing parameter: the Reynolds number Re = Um D/ν, with ν the

† Email address for correspondence: [email protected]

Downloaded from http:/www.cambridge.org/core. University of Georgia Libraries, on 03 Jan 2017 at 13:40:49, subject to the Cambridge Core terms of use, available at
http:/www.cambridge.org/core/terms. http://dx.doi.org/10.1017/jfm.2016.319
 372 J. P. Gallardo, H. I. Andersson and B. Pettersen
kinematic viscosity. Additionally, the Stokes number β = Re/KC = D2 /Tν (Sarpkaya
2005) is used by convention in combination with KC to map the different flow
regimes. A rich set of flow patterns is obtained by varying KC and β; for the
simplest cylindrical geometry, the circular cylinder, they have been well documented
by e.g. Tatsuno & Bearman (1990) and Elston, Blackburn & Sheridan (2006).
For low values of KC and β the unsteady flow remains attached to the cylinder,
and is completely symmetrical and two-dimensional, with a reflection symmetry about
the axis of oscillation. As the governing parameters are gradually increased a first
three-dimensional instability appears (Honji 1981), but a spatio-temporal symmetry
is retained in the plane normal to the axis of the cylinder. Vortex shedding and
separation are characteristics of the regimes at high KC; Williamson (1985) associated
the different shedding regimes with a certain KC range and Bearman et al. (1985)
studied the contribution from vortex shedding and separation to the in-line forces.

1.2. Primary and secondary instabilities

Contrary to the steady flow past a circular cylinder, whose primary instability is
two-dimensional, the primary instability of the oscillating flow past a circular
cylinder is three-dimensional. Previous investigators have claimed this to be a
centrifugal instability induced by the curvature of the cylinder (Honji 1981; Hall 1984).
Observations of the streaked flow by Honji were validated by instability analysis of
Hall (1984), who found a critical line in the (KC, β)-plane for this instability, in very
good agreement with Honji’s experimental results. The mushroom-shaped structures
were also visualized in experiments by Sarpkaya (1986), who extended the study
of this primary instability using several high-speed video cameras and laser-induced
fluorescence (Sarpkaya 2002). This last study resulted in a classification of the region
around the instability line into four different states: stable, marginal, unstable and
chaotic. Numerical simulations have contributed considerably to the knowledge about
the Honji instability by means of detailed visualizations of the structure of the Honji
vortices, and explanations of the dynamics involved in their formation (see e.g. Zhang
& Dalton 1999; An, Cheng & Zhao 2011; Suthon & Dalton 2011, 2012).
As the oscillation amplitude Um increases further, vortices are formed and shed
during each half-cycle. This results in secondary instabilities that break the symmetry
in the plane normal to the cylinder axis, producing a set of fascinating flow patterns.
Eight different regimes were visualized and grouped according to their characteristics
by Tatsuno & Bearman (1990) at 1.6 6 KC 6 15 and 5 6 β 6 160. This classification
was extended by Elston et al. (2006), who used Floquet analysis and direct numerical
simulations to map the symmetry-breaking instabilities in the (KC, β)-plane at
KC 6 10 and β 6 100. Depending on the Stokes number, two different manifestations
of the secondary instability were observed in their study: at low Stokes numbers the
instability was synchronous with the imposed oscillation, resulting in a characteristic
boomerang-shaped mode, whereas higher Stokes numbers resulted in a quasi-periodic
instability with a clear second period.

1.3. Instability theory for elliptic cylinders

Hall (1984) extended his instability theory for circular cylinders to more general
geometries; in particular, cylinders with an elliptic cross-section. He showed that for
certain orientations of an elliptic cylinder the locations at which the instability arises
depend on the aspect ratio of the cylinder, i.e. the ratio between the minor and major
Downloaded from http:/www.cambridge.org/core. University of Georgia Libraries, on 03 Jan 2017 at 13:40:49, subject to the Cambridge Core terms of use, available at
http:/www.cambridge.org/core/terms. http://dx.doi.org/10.1017/jfm.2016.319
 Oscillatory flow past elliptic cylinders 373
axis Λ = b/a. Throughout the present paper we adopt the parametric representation
of the ellipse,
a b
x= cos φ and y= sin φ, (1.1a,b)
2 2
with 0 6 φ 6 2π. The most stable configuration is that with the major axis aligned
with the oscillatory flow direction, in which case the orientation angle α is zero.
For a circular cylinder (Λ = 1) the instability is located at an angle φ1 = π/2 from
the stagnation points (i.e. at the shoulders of the cylinder), which is the region where
the flow reaches maximum acceleration. According to Hall, for an elliptic cylinder
with α = 0 the location of the
√ instability remains at the shoulders of the cylinder for Λ
larger than a critical value 3/5. As Λ decreases below this critical value the location
of the instability is displaced towards the stagnation points. In this case six critical
points arise on the ellipse: two at ±π/2 from the stagnation points, and four at a
critical angle φm symmetrically located with respect to the minor and major axes. As
pointed out by Hall (1984), this shift in the critical points is due to the increasing
eccentricity of the ellipse as the aspect ratio Λ decreases. From the stability analysis
he showed that the flow is neutrally stable at an angle φm by considering the maxima
of the function

S(φ) = sin2 (φ − α)(sin2 φ + Λ2 cos2 φ)−5/2 . (1.2)

√
When, for instance, α = 0 and Λ < 3/5 two additional critical points appear at
1/2
2Λ2

−1
φ = φ2 = sin and φ = φ3 = π − φ2 , (1.3a,b)
3(1 − Λ2 )
and the instability curve depicted in figure√1(b) bifurcates into two branches for aspect
ratios Λ below the critical aspect ratio 3/5. For lower aspect ratios the angle φm
departs from the standard value π/2 for circular cylinders.
Whether the oscillating flow is stable or unstable is determined by the local Taylor
number
2 1/2 2 Λ1/2 (1 + Λ)2 sin2 φ
Ta` = β KC , (1.4)
π3/2 (sin2 φ + Λ2 cos2 φ)5/2
which represents a measure of the ratio between centrifugal and viscous forces. Hall’s
(1984) linear stability analysis demonstrated that the flow is neutrally stable when
Ta` = 11.99 at φ = φm and accordingly unstable for higher values of Ta` . For the sake
of convenience, Hall also introduced the global or non-local Taylor number
2 1/2 2 −1/2
Ta = β KC Λ . (1.5)
π3/2
Hall’s instability diagram for the particular orientation α = 0◦ of the elliptic cylinder
has been reproduced in figure 1(a) in which the curves for neutral instability are
shown. Here, Ta0 = 11.99/4 ≈ 3.0 is the critical value of the global Taylor number
Ta for Λ = 1, i.e. for a circular cylinder.
Apart from Hall’s analysis, there are not many studies involving oscillating flows
past an elliptic cylinder. Badr & Kocabiyik (1997) investigated the oscillating
viscous flow around an elliptic cylinder by integrating the vorticity-transport
Downloaded from http:/www.cambridge.org/core. University of Georgia Libraries, on 03 Jan 2017 at 13:40:49, subject to the Cambridge Core terms of use, available at
http:/www.cambridge.org/core/terms. http://dx.doi.org/10.1017/jfm.2016.319
 374 J. P. Gallardo, H. I. Andersson and B. Pettersen

(a) 8 (b)
7
6
5
4
3
2
1
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

F IGURE 1. Instability curves reproduced in accordance with Hall’s (1984) analysis for an
orientation angle α = 0◦ . (a) Cusp-shaped curve depicting the dependence of the critical
Taylor number Ta/Ta0 on Λ; and√(b) dependence of the critical angle φm on Λ. The
critical case corresponds to Λ = 3/5 ≈ 0.775 and the lowest-eccentricity case is for
Λ = 1.0 (circular cylinder).

equation in combination with the stream function equation. However, their study
was two-dimensional and with focus on separated flow. In the course of the final
revision of this paper, we incidentally became aware of another computational study
of oscillatory flow around a cylinder with an elliptic cross-section by Yang (2014).
The three-dimensional Navier–Stokes equations were integrated numerically with the
same finite-element method as that already used by An et al. (2011) and Yang et al.
(2014) with the view to explore the effects of the elliptic cross-sections.

1.4. Objectives and approach

Although the oscillating flow problem has been extensively studied for circular
cylinders, little is known about the different regimes and instabilities resulting from
the oscillating flow past an elliptic cylinder. One may intuitively anticipate that the
unsteady flow field in the vicinity of an elliptic cylinder crucially depends on its
eccentricity and only resembles that around a circular cylinder when Λ ≈ 1. This
problem has practical importance for some applications such as piles, submarine
sections and wings. The main challenge in this case is the presence of two additional
parameters: the aspect ratio Λ and the orientation α of the ellipse relative to the
flow direction. In the present study we investigate, for the first time to the authors’
knowledge, the primary three-dimensional instability for a cylinder with an elliptic
cross-section oriented with its major axis parallel to the main flow direction (α = 0◦ );
this orientation, according to Hall’s analysis, exhibits a cusp-shaped instability curve.
Four representative aspect ratios are chosen for this purpose. In order to obtain
sufficient details of the flow dynamics and the three-dimensional structure of the
primary instability, our approach is to solve directly the full three-dimensional
Navier–Stokes equations. Although the main focus here is on the primary instability
predicted by Hall (1984), also known as the Honji instability, we also present results
of the transition towards the secondary instability that motivate future work.
This paper is arranged in the following way. First, in § 2 we define the problem
according to the governing parameters discussed above, and describe the flow
configuration used for the simulations. Next, in § 3 we present details concerning
Downloaded from http:/www.cambridge.org/core. University of Georgia Libraries, on 03 Jan 2017 at 13:40:49, subject to the Cambridge Core terms of use, available at
http:/www.cambridge.org/core/terms. http://dx.doi.org/10.1017/jfm.2016.319
 Oscillatory flow past elliptic cylinders 375

0.5

–0.5
–1.5 –1.0 –0.5 0 0.5 1.0 1.5

F IGURE 2. Sketch of the different cylindrical configurations considered in the present

study. The orientation angle is α = 0◦ , which corresponds to the most stable configuration
according to Hall (1984). Here√ the selected ratios between the minor and major axis
Λ = b/a are 1/3, 1/2, 2/3, 3/5; the circular cylinder with Λ = 1 is also included as
reference. An important parameter used throughout the present study is the angle φ, which
is measured anticlockwise from the stagnation point (a/2, 0).

the numerical methodology and the grid studies. The presentation of results starts
with a preliminary exploration of the effects of ellipticity in § 4, where we compare
the vortical structures for different Λ at the same phase. This is followed by a
detailed analysis of the evolution of the flow structures for the case with highest
ellipticity (Λ = 1/3) in § 5. In this particular cylinder case we consider five different
values of the parameter β for fixed KC. Finally, a summary and concluding remarks
are given in § 6.

2. Problem definition and methodology

2.1. Governing parameters
Hall’s instability theory for an elliptic cylinder is valid for any orientation α
and aspect ratio Λ. Due to the high computational cost of directly solving the
three-dimensional Navier–Stokes equations, we have chosen to study only flow
configurations with orientation α = 0◦ (i.e. with the major axis of the ellipse √
aligned with the oscillatory flow), and aspect ratios Λ = 1/3, 1/2, 2/3 and 3/5
(see figure 2).
According to Hall’s theory α = 0◦ is the most stable configuration and exhibits
a cusp-shaped instability curve, thereby providing the opportunity to observe
three-dimensional instabilities for Taylor numbers above the lower branch of this
cusp-shaped curve in figure 1(a). For carefully selected combinations of the governing
parameters KC and β, we expect to observe vortices similar to those observed by
Honji (1981) in his experiments with an oscillating circular cylinder. Given the
direction of the oscillating flow, the Reynolds number Re is defined using the
transverse scale b, whereas the Keulegan–Carpenter number KC is defined with the
major axis a as the length scale,
Um T Um b Re ab
KC = , Re = , β= = . (2.1a−c)
a ν KC νT
Thus, with basis in Hall’s theory, several combinations of KC and β were considered
for the simulations, and the cases with clear three-dimensional instabilities were
selected for further examination throughout this paper. Table 1 presents a summary
of the main parameters for these cases. In table 1 we also include an estimate for the
Downloaded from http:/www.cambridge.org/core. University of Georgia Libraries, on 03 Jan 2017 at 13:40:49, subject to the Cambridge Core terms of use, available at
http:/www.cambridge.org/core/terms. http://dx.doi.org/10.1017/jfm.2016.319
 376 J. P. Gallardo, H. I. Andersson and B. Pettersen

Case Λ KC β Re Ta λz /b λ0z /b
C1a 2.0 200 400 35.19 — —
C1b 2.0 300 600 43.10 1.00 0.70
C1c 1/3 2.0 400 800 49.77 1.00 0.60
C1d 2.0 500 1000 55.64 ≈1.00 0.54
C1e 2.0 600 1200 60.95 ≈1.00 0.49
C2 1/2 2.0 300 600 35.19 0.80–1.00 0.57
C3 2/3 2.0 200 400 24.88 0.80 0.60
√
C4 3/5 2.0 200 400 23.09 0.80 0.56
C5 1 2.0 200 400 20.32 0.67 0.49
TABLE 1. Summary of the cases exhibiting three-dimensional instabilities considered in
the present study. Here KC is the Keulegan–Carpenter number, β the Stokes number, Re
the Reynolds number, Ta the global Taylor number, λz the axial wavelength of the three-
dimensional instability obtained from the simulations and λ0z the theoretical estimate (2.3)
for this axial wavelength. Note that the critical value of the global Taylor number for a
circular cross-section is Ta0 = 11.99/4 ≈ 3.0.

axial wavelength of the three-dimensional instability λ0z . This theoretical prediction

is obtained from the scaling in Hall’s equations (2.1b) and (1.1a), together with his
critical wavenumber kc = 0.51 (Hall’s equation (3.1b)),
2π νT 1/2
 
0
λz = . (2.2)
kc π
Using the Stokes number β as defined in (2.1) we obtain
2π
λ0z = (πΛβ)−1/2 b. (2.3)
kc
According to expression (2.3), λ0z decreases as Λ and β increase. Although the
predicted values λ0z in table 1 are always lower than those obtained from the
numerical simulations, predicted and observed axial wavelengths have the same
order of magnitude.

2.2. Governing equations

The governing equations are the incompressible Navier–Stokes equations for a
fluid with constant density ρ and kinematic viscosity ν, which expressed in
non-dimensional form are
∂ui
= 0, (2.4)
∂xi
1 ∂ui ∂ui ∂p 1 ∂ 2 ui
+ uj =− + . (2.5)
KC ∂t ∂xj ∂xi Re ∂xj ∂xj
The flow field is determined by the velocity ui in each of the three Cartesian directions
xi , with i = 1, 2, 3, and the pressure p. In the present context u, v, w refer to the
streamwise, cross-stream and spanwise velocity components, respectively, and t is non-
dimensional time normalized with T.
Downloaded from http:/www.cambridge.org/core. University of Georgia Libraries, on 03 Jan 2017 at 13:40:49, subject to the Cambridge Core terms of use, available at
http:/www.cambridge.org/core/terms. http://dx.doi.org/10.1017/jfm.2016.319
 Oscillatory flow past elliptic cylinders 377
3. Numerical aspects
3.1. Numerical solution of the governing equations
The flow field is computed numerically by directly solving the governing momentum
equation (2.5) subject to the incompressibility constraint (2.4). To accomplish this, we
use the Navier–Stokes solver MGLET (Manhart, Tremblay & Friedrich 2001; Manhart
2004). This is a finite-volume code with second-order spatial accuracy in which the
flow variables are defined on a staggered Cartesian grid. Good convergence properties
are achieved in the numerical solution of the Poisson equation for pressure correction
by using the strongly implicit procedure (SIP) by Stone (1968), accelerated with a
multigrid algorithm. For the time integration we use an explicit low-storage third-order
Runge–Kutta scheme (Williamson 1980).
To represent the elliptic cylinders inside the computational domain, we use an
immersed boundary (IB) method based on direct forcing to specify the no-slip and
impermeability boundary conditions at the cylinder surface. In order to ensure mass
conservation a flux correction based on the cell volumes at the fluid–solid interface is
applied. The description of this IB method and its validation can be found in Peller
et al. (2006).
The computational domain is defined as a box in which the cylinder axis coincides
with the origin in the (x, y)-plane. In the (x, y)-plane the computational box spans
50b × 50b. This size was chosen based on previous experience with oscillatory flow
around a circular cylinder, and domain sizes used by previous authors for similar flows
(see e.g. Elston et al. 2006; An et al. 2011). With the exception of cases C1d and
C1e , vortices are not shed from the cylinder, and when shedding occurs they never
reach the boundaries of the domain. Depending on the spanwise wavelength λz of
the primary instability two different spanwise-domain lengths Lz of 4b and 8b were
considered.
Concerning numerical boundary conditions, we use the following:
(i) Oscillating velocity [u, 0, 0] = Um [cos(2πt/T), 0, 0], with Um the maximum
velocity amplitude at the (y, z)-plane x/b = −25.
(ii) At the (y, z)-plane at x/b = 25 a Neumann boundary condition is imposed for the
velocities (∂u/∂x = ∂v/∂x = ∂w/∂x = 0), and the pressure is set to zero (p = 0).
(iii) Periodicity is prescribed at the top and bottom (x, y)-planes to mimic an infinitely
long cylinder.
(iv) At the remaining side (x, z)-planes we prescribe a free-slip boundary condition.
This corresponds to v = 0 and ∂u/∂y = ∂w/∂y = 0 at y/b = ±25.

3.2. Grid refinement tests

Grid tests were conducted for the cases with Λ = 1/3 because, as mentioned above,
these cases exhibit the largest Re. The main quantities considered to study the
different grid resolutions were the in-line force coefficients, and the growth of the
instabilities based on the spanwise velocity w. Force coefficients were obtained by
decomposing the drag and inertia forces according to Morison’s equation (Morison
et al. 1950). Since the KC numbers considered in this study are relatively low, the
inertia force dominates over the drag force in all the cases considered. Three different
grid resolutions for case C1c were examined (see table 2). In all the simulations the
time step 1t was set to one-tenth of ∆min /Um , with ∆min the minimum grid spacing,
in order to ensure low divergence and numerical stability.
Downloaded from http:/www.cambridge.org/core. University of Georgia Libraries, on 03 Jan 2017 at 13:40:49, subject to the Cambridge Core terms of use, available at
http:/www.cambridge.org/core/terms. http://dx.doi.org/10.1017/jfm.2016.319
 378 J. P. Gallardo, H. I. Andersson and B. Pettersen

10
8
5 7
6
0
5
–5 19.7 19.8

–10
10 11 12 13 14 15 16 17 18 19 20

F IGURE 3. The time evolution of the in-line force coefficient CFx = Fx /0.5ρUm2 Lz b for
three different grid resolutions at Λ = 1/3 (C1c ): fine (——), medium (- - -) and coarse
(· · · · · ·) grids. The inset plot zooms into the peaks of the force coefficients, depicting
minute differences for different grid resolutions.

Case Grid ∆min /b Nx × Ny × Nz Cd Cm Growth rate

Coarse 0.010 544 × 352 × 64 0.7303 1.4248 0.0009
C1c Medium 0.0075 674 × 400 × 64 0.7238 1.4231 0.0140
Fine 0.005 896 × 512 × 64 0.7256 1.4298 0.0123
TABLE 2. The drag and inertia coefficients, Cd and Cm respectively, and linear growth
rates for Λ = 1/3 and Re = 800 for different grid resolutions.

Table 1 shows that the wavelength λz /b for case C1c is 1.0, and hence a grid spacing
in the spanwise direction ∆z /b = 0.0625 was sufficient to resolve the three-dimensional
flow structures.
Since the excursions from Re = 800 are relatively low for C1d and C1e , these cases
were simulated with the same grid resolutions in the x- and y-directions as for case
C1c . All the cases with Λ = 1/3 were simulated using a spanwise length Lz /b = 8. The
wavelength of the three-dimensional instability for cases C1b and C1c is comparatively
large, therefore a spanwise length of Lz /b = 8 was considered sufficient. However, no
differences in the flow properties were observed for this case with a shorter spanwise
length.
Table 2 shows that the force coefficients Cd and Cm exhibit only a modest
dependence on the grid refinement, with differences only in the third significant
digit. The non-monotonic trend from the coarse to fine grids for case C1c can be
explained by the fact that the three-dimensional instabilities are not triggered in the
coarse grid. Figure 3 shows the time evolution of the in-line force coefficients for
case C1c with different grid resolutions. Even in the inset plot zooming into the peaks
the differences are barely distinguishable for the different resolutions considered.
When studying the primary three-dimensional instability, an important factor in
deciding an adequate grid resolution is the development of the instability in time. This
issue was examined for case C1c in figure 4, where the time history of the spanwise
velocity w is shown for different grid resolutions. This velocity component was
sampled at regularly spaced points along a line parallel to the cylinder axis, located
at φ = 48◦ from the stagnation point (−a/2, 0), and at a distance of 0.15b from the
cylinder surface. No artificial perturbations were added to trigger the instability. It
is evident from figure 4 that the growth of this velocity component converges to a
rate of the order of 0.01 with sufficient grid resolution (see also table 2). Even more
Downloaded from http:/www.cambridge.org/core. University of Georgia Libraries, on 03 Jan 2017 at 13:40:49, subject to the Cambridge Core terms of use, available at
http:/www.cambridge.org/core/terms. http://dx.doi.org/10.1017/jfm.2016.319
 Oscillatory flow past elliptic cylinders 379

(a) 0.10 (b) 0.10

0.05 0.05
0 0

–0.05 –0.05

–0.10 –0.10
5 10 15 20 25 30 5 10 15 20 25 30

(c) 0.10
0.05
0
–0.05
–0.10
5 10 15 20 25 30

F IGURE 4. Time history of the spanwise velocity w depicting the growth of the three-
dimensional instability for different grid resolutions. Case considered is C1c with Λ = 1/3,
KC = 2 and β = 400; (a) coarse, (b) medium and (c) fine grid resolutions. The dashed
curve in (a) reflects the time variation of the free-stream velocity.

striking is the almost total absence of three-dimensional instabilities after 30 cycles

using a coarse grid resolution (figure 4a). We recall that in this case the refinement
is in the (x, y)-plane only.
The growth rate is introduced as a measure of how fast the instability grows until
regular oscillations with almost constant amplitude have been established. The actual
growth rate reported in table 2 is non-dimensionalized using T and estimated from the
region in which the envelope of the oscillating signals is linear, see e.g. figure 4.
Based on the results above, we conclude that the fine grid resolutions in table 2
are sufficient to study the inherently unsteady dynamics of the three-dimensional flow
fields in the present cases.

4. Effects of ellipticity
Honji (1981) named the pattern that arose from the three-dimensional instability
streaked flow due to the characteristic bands that his flow visualizations revealed.
Later, subsequent investigations reported mushroom-like structures in the visualizations
over planes normal to the cylinder wall (see e.g. Sarpkaya 1986; Tatsuno & Bearman
1990; Zhang & Dalton 1999; An et al. 2011; Suthon & Dalton 2012). If, as
theoretically predicted by Hall (1984), the same type of centrifugal instability occurs
in the oscillating flow past elliptic cylinders, similar vortical structures consisting of
pairs of counter-rotating vortices are expected.
Figure 5 shows snapshots at phase T/4 of the instantaneous isosurfaces of
streamwise vorticity ωx for the different values of Λ considered in this study. For
Λ = 2/3 and above we observe the same distinctive mushroom-like structures reported
by previous authors (Zhang & Dalton 1999; Sarpkaya 2002; An et al. 2011; Suthon &
Dalton 2011). Staggered rows of counter-rotating vortices are seen on each side of the
cylinder. When Λ = 1 the vortices are equally sized and evenly distributed, in good
Downloaded from http:/www.cambridge.org/core. University of Georgia Libraries, on 03 Jan 2017 at 13:40:49, subject to the Cambridge Core terms of use, available at
http:/www.cambridge.org/core/terms. http://dx.doi.org/10.1017/jfm.2016.319
 380 J. P. Gallardo, H. I. Andersson and B. Pettersen

(a) (b) (c) (d ) (e)

z z z z z
x x x x x

(x, z)-view

(f) (g) (h) (i) ( j)

z z z z z
y y y y y

(y, z)-view

F IGURE 5. Instantaneous isosurfaces of streamwise vorticity ωx = ±0.5 at phase T/4 for

√
aspect ratios of Λ = 1/3 (a,f ), 1/2 (b,g), 2/3 (c,h), 3/5 (d,i) and 1 (e,j). Red isosurfaces
correspond to positive vorticity and blue to negative vorticity. Corresponding cases in
table 1 are C1c , C2 , C3 , C4 and C5 .

agreement with the observations by An et al. (2011); six pairs of counter-rotating

√
vortices at each side give a wavelength of 0.67b, whereas for Λ = 2/3 and 3/5 the
wavelength increases to 0.8b. The (y, z)-views for Λ = 2/3 and above show that tubes
on one side of the cylinder join the tubes with same vorticity sign at the opposite side
of the cylinder, with opposite vorticity isosurfaces connecting them at the stagnation
line as observed by Suthon & Dalton (2012) for Λ = 1. As Λ decreases the size of
the flow structures as well as their distribution along the span varies to some extent,
indicating that the structures are not fully stable. Although according to Hall’s theory
the instability location for Λ = 2/3 is about φ = π/4 (figure 1b), the upper and lower
branches of the instability curves in figure 1(a) are very close to each other, and the
expected effects of ellipticity are therefore modest.
When Λ = 1/2 the distribution of the flow structures is uneven, and the wavelength
is somewhere between 0.8b and 1.0b. Shifting of flow structures towards the
stagnation line on the right in the (x, z)-views indicates that the effects of ellipticity
start to be more pronounced, and agrees well with Hall’s theory. The (y, z)-view
for the lowest aspect ratio considered, Λ = 1/3, shows that the bands of vorticity
isosurfaces with opposite sign at each side of the cylinder are aligned, contrary to the
staggered rows seen for instance at Λ = 1. Next to the stagnation line the pattern of
the vortical structures for Λ = 1/3 is distinctly different from that seen at higher Λ.
The streamwise vortices for Λ = 1/3 exhibit a flattened rather than the characteristic
tubular shape of the Honji vortices. Suthon & Dalton (2011) described the formation
of the vortex pairs for a circular cylinder as spanwise vorticity ωz which undergoes
stretching and reorientation into streamwise and cross-stream vorticity. It is plausible
that this same process is responsible for the generation of the vortical structures at
Λ = 1/3, however, as Λ decreases the eccentricity of the ellipse increases, giving rise
to these characteristic spanwise vortical structures.
The isosurfaces of spanwise vorticity ωz √ in figure 6 exhibit the same rib-like
layers mentioned above for cases Λ = 2/3, 3/5 and 1, and agree well with the
visualizations reported by An et al. (2011) for circular cylinders. In addition to the
Downloaded from http:/www.cambridge.org/core. University of Georgia Libraries, on 03 Jan 2017 at 13:40:49, subject to the Cambridge Core terms of use, available at
http:/www.cambridge.org/core/terms. http://dx.doi.org/10.1017/jfm.2016.319
 Oscillatory flow past elliptic cylinders 381

(a) (b) (c) (d) (e)

z z z y z y z
y y y
x x x x
x

F IGURE 6. Perspective view of the spanwise vorticity ωz = ±1.0 at phase T/4 for aspect
√
ratios of Λ = 1/3 (a), 1/2 (b), 2/3 (c), 3/5 (d) and 1 (e). Red isosurfaces correspond
to positive vorticity and blue to negative vorticity. Corresponding cases in table 1 are C1c ,
C2 , C3 , C4 and C5 .

characteristic undulations in the layers of vorticity, we observe two deformed tubes

with their cores aligned with the axis of the cylinder for the cases with Λ = 1/3 and
1/2. The emergence of these structures at low Λ is an obvious consequence of the
increased eccentricity of the ellipses.
Contrary to the present study, Yang (2014) focussed √on modest eccentricities
Λ & 0.866, i.e. well above the critical aspect ratio 3/5 ≈ 0.775 at which
the instability diagram in figure 1(a) bifurcates. The only exceptions were the
cross-sections Λ = 0.766 and 0.5 that were studied at the relatively high β = 400.
Although Yang’s KC = 2.0 matches that used for cases C2 and C4 , the twice as high
Stokes number β makes the resulting flow fields qualitatively different.

5. Development of the flow structures for Λ = 1/3

In the following we examine results for Λ = 1/3, KC = 2, and varying β = 200,
300, 400, 500 and 600. This cylinder configuration has a relatively low aspect ratio
Λ which gives rise to peculiar vortex patterns compared to other configurations with
more modest eccentricity.
The lack of previous physical observations of the instabilities generated by
oscillating flow past an elliptic cylinder led us to conduct extensive spatio-temporal
sampling. First, the time evolution of the instability was tracked with a sampling set
consisting of 10 confocal ellipses, each with 60 sampling points equidistant at an
angle of 6◦ , yielding a total of 600 sampling points in each (x, y)-plane. In the axial
direction samples were taken at 41 equidistant (x, y)-planes separated by 0.1b, thus
the total number of sampling points is 24 600. The time windows for the sampling
varied depending on how fast the instabilities develop. For case C1a , a total of 30
cycles were sampled; for case C1b 80 cycles; and for cases C1c to C1e 60 cycles.
The time series portraying the evolution of the three-dimensionality of the flow
for cases C1b –C1e are shown in figure 7. A line parallel to the cylinder axis at the
instability location φ2 given by (1.3) and at a distance from the cylinder of 0.2b in
the wall-normal direction was chosen to show the temporal and spatial development
of the spanwise velocity w. The emergence of this velocity component renders the
flow three-dimensional, indicating the occurrence of the primary instability predicted
by Hall (1984). At β = 200 no traces of spanwise velocity w were observed, thus the
Downloaded from http:/www.cambridge.org/core. University of Georgia Libraries, on 03 Jan 2017 at 13:40:49, subject to the Cambridge Core terms of use, available at
http:/www.cambridge.org/core/terms. http://dx.doi.org/10.1017/jfm.2016.319
 382 J. P. Gallardo, H. I. Andersson and B. Pettersen

(a) 4 (b) 4
3 3
2 2
1 1
0 I II 0 I II III
–1 –1
–2 –2 0.10
–3 –3 0.08
–4 –4 0.06
40 45 50 55 60 65 70 75 80 5 10 15 20 25 30 35 40 45 50 55 60 0.04
0.02
0
(c) 4 (d) 4 –0.02
3 3 –0.04
2 2 –0.06
1 –0.08
1
–0.10
0 0
–1 –1
–2 –2
–3 –3
–4 –4
5 10 15 20 25 30 35 40 45 50 55 60 5 10 15 20 25 30 35 40 45 50 55 60

F IGURE 7. Time evolution of the spanwise velocity component w for Λ = 1/3 and KC = 2.
Roman numerals I, II and III denote the different phases identified by Yang et al. (2014)
for the Honji instability on a circular cylinder. The boxes in (c,d) demarcate the regions
for sampling during five periods (figures 20 and 21). (a) β = 300; (b) β = 400, with the
dashed line indicating time for the sampling during one period in figure 18; (c) β = 500;
and (d) β = 600.

flow is two-dimensional. For β = 300 and above, characteristic strips similar to those
reported by An et al. (2011) for a circular cylinder are observed. This streaked flow
pattern resembles that first observed by Honji (1981) in the flow around an oscillating
circular cylinder, which gives rise to pairs of counter-rotating vortices or mushroom-
shaped coherent structures (see e.g. Sarpkaya 2002; Suthon & Dalton 2011). However,
since the bluff body involved in the present study is an elliptic cylinder, we avoid
denoting this a Honji instability.
Case C1b (figure 7a) starts to exhibit three-dimensionality after approximately 55
oscillation cycles, and the spanwise velocity w grows at a relatively slow rate of
0.0031. Two phases I and II can be distinguished in this case, according to the
classification previously presented by Yang et al. (2014). Phase I is characterized
by the absence of three-dimensional instabilities, as is the case for C1a , whereas in
phase II the instabilities grow and develop into a stable structure; essentially after
approximately 75 cycles for β = 300. When β increases to 400 (figure 7b) the
instability grows at a faster rate 0.0123. Phase I lasts about 17 cycles for this case,
and the strips develop into a stable structure thereafter (phase II). However, a third
phase III characterized by rearrangements of the strips is observed in this case after
approximately 40 cycles. A fundamental difference with the instability of the flow
past a circular cylinder (and elliptic cylinders with low ellipticity) is that the Honji
vortices are sustained throughout one period of oscillation (Sarpkaya 2002; An et al.
2011; Suthon & Dalton 2011), whereas the three-dimensional instability in the present
case with Λ = 1/3 grows and decays periodically.
Downloaded from http:/www.cambridge.org/core. University of Georgia Libraries, on 03 Jan 2017 at 13:40:49, subject to the Cambridge Core terms of use, available at
http:/www.cambridge.org/core/terms. http://dx.doi.org/10.1017/jfm.2016.319
 Oscillatory flow past elliptic cylinders 383

0.05

0.04

Linear growth rate

0.03

0.02

0.01

300 400 500 600

F IGURE 8. Estimated growth rates for the Stokes numbers considered in the simulations
with KC = 2 and Λ = 1/3.

Increasing β further for a given KC leads to complex interactions between

the flow structures. Case C1d in figure 7(c) exhibits an early development of the
three-dimensional flow (≈10 cycles) with a linear growth rate of 0.0277, and then
stable structures, depicted by the parallel strips, for approximately five cycles. This is
followed by a further development into more complex flow involving rearrangements
along the span of the cylinder. A similar development of the flow structures has
been reported by An et al. (2011) for circular cylinders, which involves merging
and annihilation of the vortices. For a circular cylinder, experimental observations by
Honji (1981) and Sarpkaya (2006) report transition to turbulence when KC = 2 and
β = 580. This was confirmed by the direct numerical simulations of An et al. (2011)
for KC = 2 and β = 600, where an irregular flow pattern and abrupt changes in the
flow structures over time intervals as short as one cycle were reported. In figure 7(d),
the same (KC, β)-combination of parameters for an elliptic cylinder with aspect ratio
Λ = 1/3 gives rise to an irregular flow pattern and more frequent reorganizations
of the strips than at β = 500. The linear growth rate during the initial development
of the three-dimensional flow is 0.0472. Despite this apparently chaotic behaviour,
the flow still retains the basic paired structure observed at lower β and it cannot be
classified as turbulent.
The estimated growth rates for the different Stokes numbers are shown in figure 8.
For this particular aspect ratio (Λ = 1/3) and KC = 2, the growth rate increases
monotonically with β, i.e. as the stabilizing influence of the fluid viscosity is reduced.
Moreover, by extrapolating the growth rate to zero using a second-order polynomial,
the critical β-value ≈247.4 for the onset of the instability is obtained.
Further information about the time evolution of the instability can be obtained by
plotting the velocity traces at a particular spanwise location. This allows us to compare
the strength of the three-dimensional flow for different values of β, and to examine the
periodicity of the flow at different locations. In figure 9 two characteristic spanwise
locations were selected for each case in order to visualize the traces of w during five
cycles; the two lines aim to cut the positive and negative part of the strips, and for
β = 400 the time window is located within phase II. In addition to the data at the
primary instability location φ2 , traces at φ1 = π/2 corresponding to the upper branch
of the cusp-shaped curve shown in figure 1(a) are also included. We recall that for
Downloaded from http:/www.cambridge.org/core. University of Georgia Libraries, on 03 Jan 2017 at 13:40:49, subject to the Cambridge Core terms of use, available at
http:/www.cambridge.org/core/terms. http://dx.doi.org/10.1017/jfm.2016.319
 384 J. P. Gallardo, H. I. Andersson and B. Pettersen

(a) 0.2 (b) 0.2

0.1 0.1
0 0
–0.1 –0.1
–0.2 –0.2
75 76 77 78 79 80 75 76 77 78 79 80
(c) 0.2 (d ) 0.2
0.1 0.1
0 0
–0.1 –0.1
–0.2 –0.2
25 26 27 28 29 30 25 26 27 28 29 30
(e) 0.2 ( f ) 0.2
0.1 0.1
0 0
–0.1 –0.1
–0.2 –0.2
35 36 37 38 39 40 35 36 37 38 39 40
(g) 0.2 (h) 0.2
0.1 0.1
0 0
–0.1 –0.1
–0.2 –0.2
30 31 32 33 34 35 30 31 32 33 34 35

F IGURE 9. Time history of the spanwise velocity component w over five cycles for
Λ = 1/3, KC = 2 and β = 300 (a,b), 400 (c,d), 500 (e,f ) and 600 (g,h). Two samples
inside the strip pairs shown in figure 7 are selected, and samples are taken at 0.2b from
the cylinder and at critical angles φ2 = 16.8◦ (a,c,e,g) and φ1 = 90◦ (b,d,f,h). Samples
were aimed to trace the two poles of a vortex pair, measuring positive (black lines)
and negative (grey lines) velocities. The dashed curves indicate the time variation of the
imposed oscillating flow U.

√
a circular cylinder (Λ = 1) and elliptic cylinders with Λ > 3/5 this is the location
where the Honji instability is initialized. In addition, the dashed curves in figure 9
depict the time variation of the imposed oscillations.
The spanwise velocity traces when β = 300 and 400 at φ2 exhibit a main peak
of about 5 % the free-stream velocity preceded by a secondary peak of lower
magnitude. These peaks are synchronized with the ambient flow with a phase shift
of approximately π/4. A similar behaviour is observed at β = 500 and 600, but
the traces are more irregular with amplitudes of approximately 10 % the incoming
flow in this case. Overall, the three-dimensional instability at this location reaches its
maximum intensity during flow reversal at times t = T/4 and then decays to zero at
≈T/2. At φ1 we observe two peaks within a cycle which are phase aligned with the
ambient flow. The intensity of the three-dimensional flow grows from ≈5 % of the
velocity Um at β = 300 to approximately 20 % of Um at β = 600. Indeed, for β = 600
the intensity of w at φ1 exceeds that at φ2 , suggesting a transition in the primary
instability location from (φ2 , φ3 ) to φ1 (see figure 1).
Downloaded from http:/www.cambridge.org/core. University of Georgia Libraries, on 03 Jan 2017 at 13:40:49, subject to the Cambridge Core terms of use, available at
http:/www.cambridge.org/core/terms. http://dx.doi.org/10.1017/jfm.2016.319
 Oscillatory flow past elliptic cylinders 385

(a) (b)
T 0.10 T 0.10
0.05 0.05
0 0
–0.05 –0.05

0 –0.10 0 –0.10

(c) T 0.10 (d ) T 0.10

0.05 0.05
0 0
–0.05 –0.05

0 –0.10 0 –0.10

F IGURE 10. Temporal evolution of the spanwise velocity w during one cycle along the
upper half of the elliptical cross-section for Λ = 1/3, KC = 2 and β = 300 (a), 400 (b),
500 (c) and 600 (d). The samples are taken at a wall-normal distance of 0.1b, and white
dashed lines denote the critical angles φ2 = 16.8◦ and φ3 = 163.2◦ .

An interesting result from Hall’s instability analysis is the cusp-shaped curve for
some orientations of an elliptic cylinder (see figure 1a), and the gradual displacement
of the instability points from the shoulders of the cylinder towards the stagnation
points as Λ decreases (see figure 1b). In order to validate Hall’s theory, we sampled
the spanwise velocity w around the elliptic cross-section at a normal distance of 0.1b,
where the boundary-layer velocities are around their maxima. The dashed lines in
figure 10 denote the critical angles φ2 = 16.8◦ and φ3 = 163.2◦ given by Hall’s theory
for the primary instability, i.e. (1.3). The spatio-temporal evolution of the spanwise
velocity w has been tracked in this way throughout one period T for β = 300–600 at
the spanwise locations z indicated in the plots.
Figure 10 shows similar w-evolution for β = 300, 400 and 500. The instability is
triggered around φ2 at phase T/4, and then the initial eruption of axial velocity is
pushed around the elliptic cylinder for approximately half a cycle, passing through the
secondary instability location φ1 during this process. Emergence of the axial velocity
around φ3 occurs around phase 3T/4 and then moves in the opposite direction (left)
towards φ2 . This swinging displacement of the instability along the walls of the
elliptical cross-section explains the two peaks in one cycle at φ1 seen in figure 9.
For β = 600 (figure 10d) the evolution of w-bursts along the wall of the elliptic
cylinder exhibits similar properties as those observed at lower β. However the
magnitude of the spanwise velocity in this case is more pronounced. Moreover, the
symmetry exhibited at lower β is absent in this high Stokes number case due to the
dramatic changes in the flow structures depicted by figure 7(d).
The overall agreement with Hall’s theory is very good indeed when it comes to the
prediction of the eruptions of three-dimensional flow at φ2 and φ3 . Additionally the
results presented in figure 10 show the life of this primary instability throughout one
time cycle T and its initiation at phase T/4, which corresponds to flow reversal.
Next we explore the properties of the flow in cross-sectional (x, y)-planes during a
half-cycle for β = 200–600; all the snapshots are taken at the same spanwise position
Downloaded from http:/www.cambridge.org/core. University of Georgia Libraries, on 03 Jan 2017 at 13:40:49, subject to the Cambridge Core terms of use, available at
http:/www.cambridge.org/core/terms. http://dx.doi.org/10.1017/jfm.2016.319
 386 J. P. Gallardo, H. I. Andersson and B. Pettersen
(a) 1.0 (b) 1.0
0.5 0.5
0 0
–0.5 –0.5
–1.0 –1.0
–2 –1 0 1 2 –2 –1 0 1 2

(c) 1.0 (d ) 1.0

0.5 0.5
0 0
–0.5 –0.5
–1.0 –1.0
–2 –1 0 1 2 –2 –1 0 1 2

(e) 1.0
0.5
0
–0.5
–1.0
–2 –1 0 1 2

F IGURE 11. (Colour online) Contours of normalized spanwise vorticity ωz (black, positive,
and grey, negative values) and the second largest eigenvalue λ2 (thick contours to denote
vortex cores) during half a period T/2 at intervals of T/8. In all the plots |ωz b/Um | = 10
with 1ωz b/Um = 1, and λ2 = −2. Snapshots taken at z/b = 0.4. Case C1a with Λ = 1/3,
KC = 2 and β = 200.

z/b = 0.4. At low β the flow is symmetrical with respect to the (x, z)-plane and
essentially two-dimensional along the cylinder axis. This is evident in figure 11,
in which contour lines of ωz at β = 200 are plotted together with contours of the
parameter λ2 to identify vortex cores (Jeong & Hussain 1995); black lines indicate
positive vorticity and grey lines negative vorticity. The parameter λ2 represents the
location of vortex cores, and is defined as the second largest eigenvalue of the
tensor S 2 + Ω 2 , where S and Ω are the symmetric and antisymmetric parts of the
velocity-gradient tensor, respectively. Contours of vorticity allow identification of the
Stokes layer of opposite vorticity, clearly seen at t0 + 0.125T and t0 + 0.25T in
figure 11. Two pairs of vortices with opposite signs develop symmetrically during a
half-cycle close to the stagnation points. A similar development of symmetrical pairs
of vortices has been reported for a circular cylinder (see e.g. Tatsuno & Bearman
1990; Zhang & Dalton 1999). These regions of concentrated vorticity within the pair
of vortices are convected along the cylinder by the ambient flow, however they die
out before reaching the opposite end of the ellipse as they mix with vorticity of
opposite sign in the cylinder boundary layer (figure 11 at phases t0 and t0 + 0.5T).
Apart from these vortex pairs there is no evidence of massive separation involving
vortex shedding and large excursions from the wall.
An interesting aspect of the flow around elliptic cylinders is the change in the shape
of the bluff body as the parameter Λ decreases (see figure 2), since this includes the
problems of the circular cylinder (Λ = 1) and a flat plate (Λ = 0) as limiting cases.
Downloaded from http:/www.cambridge.org/core. University of Georgia Libraries, on 03 Jan 2017 at 13:40:49, subject to the Cambridge Core terms of use, available at
http:/www.cambridge.org/core/terms. http://dx.doi.org/10.1017/jfm.2016.319
 Oscillatory flow past elliptic cylinders 387

(a) 0.75 (b) 0.75 (c) 0.75

0.70 0.70 0.70
0.65 0.65 0.65
0.60 0.60 0.60
0.55 0.55 0.55
0.50 0.50 0.50
–1 0 1 –1 0 1 –1 0 1

(d ) 0.75 (e) 0.75

0.70 0.70
0.65 0.65
0.60 0.60
0.55 0.55
0.50 0.50
–1 0 1 –1 0 1

F IGURE 12. Velocity profiles during half a period taken at φ2 = π/2. Case with Λ = 1/3,
KC = 2 and β = 200 (◦ ◦ ◦) and Stokes’ theoretical solution (5.1) for oscillating flow over
a fixed wall (−).

Stokes provided the first theoretical solution of the sinusoidal oscillatory motion of
unbounded fluid over an infinite flat plate, which is known as Stokes’ second problem.
Although still far from being a flat plate, an elliptic cross-section with aspect ratio
Λ = 1/3 differs considerably from a circular cross-section, therefore a comparison with
Stokes’ theoretical solution is meaningful in the present context.
The solution for oscillating flow over a fixed and infinitely long wall (White 2006)
with free-stream velocity Um cos(2πt/T) is given by
u(y, t)  t  t 
= cos 2π − exp(−η) cos 2π − η , (5.1)
Um T T
√
with η = πβy/b. The thickness δ of the oscillating boundary layer when exp(−η) =
0.01, i.e. η ≈ 4.6, is
δ 4.6
≈√ . (5.2)
b πβ
Figure 12 shows the velocity profiles for β = 200 taken at the upper shoulder of the
ellipse (φ = π/2) together with Stokes’ theoretical solution; the profiles are plotted
at the same phases as those in figure 11. Clearly, at a distance of 0.05b from the
wall the agreement between the simulation and the theory is very good. However,
the overshooting higher up in the boundary layer is always larger in the simulations.
The theoretical prediction of the boundary-layer thickness for a flat plate at β = 200
yields δ/b = 0.184, differing substantially from the simulations because, in this case,
the velocity takes longer to decay due to the perturbation induced by the elliptical
cross-section. It is expected that the agreement with the theoretical predictions will
improve as the shape parameter Λ decreases. Nevertheless the vorticity attains very
low values at 0.2b from the wall due to the low gradients in the velocity profiles in
figure 12.
Downloaded from http:/www.cambridge.org/core. University of Georgia Libraries, on 03 Jan 2017 at 13:40:49, subject to the Cambridge Core terms of use, available at
http:/www.cambridge.org/core/terms. http://dx.doi.org/10.1017/jfm.2016.319
 388 J. P. Gallardo, H. I. Andersson and B. Pettersen

(a) 1.0 (b) 1.0

0.5 0.5
0 0
–0.5 –0.5
–1.0 –1.0
–2 –1 0 1 2 –2 –1 0 1 2

(c) 1.0 (d ) 1.0

0.5 0.5
0 0
–0.5 –0.5
–1.0 –1.0
–2 –1 0 1 2 –2 –1 0 1 2

(e) 1.0
0.5
0
–0.5
–1.0
–2 –1 0 1 2

F IGURE 13. (Colour online) As figure 11 but β = 300.

The velocity profiles at phases t0 and t0 + 0.5T give the impression that separation
occurs at the wall. However, the region where ωz vanishes (or equivalently the wall
shear stresses vanish) does not necessarily correspond to separation point as in steady
approaching flow. The largest difference in velocity between the simulations and the
flat-plate solution occurs at phase t0 + 0.375T, when the magnitude of free-stream
velocity reaches its maximum value Um . Here the magnitude of the velocities of the
flow around the elliptic cylinder exceeds those of the flat plate by about 0.5Um . When
it comes to the distribution of the velocities the results in figure 12 exhibit good
qualitative agreement.
From the comparisons in figure 12 we learned that the simple analytic solution (5.1)
for oscillating flow over an infinitely long flat plate compares surprisingly well with
the computed velocity profiles at φ = 90◦ on the curved surface of the elliptic cylinder.
This demonstrates not only that the effect of the local curvature of the ellipse is almost
negligible, but also that the vortices in the stagnation regions seen in figure 11 have a
negligible effect on the flow field at the shoulders for this particular aspect ratio. For
even lower aspect ratios, i.e. Λ < 1/3, an even closer correspondence between the flow
field around the ellipse and the analytical flat-plate solution is expected.
At higher β the flow in the cross-sectional plane retains the characteristics observed
at β = 200 when it concerns the generation of a counter-rotating vortex pair at each
half-cycle (see for instance figures 13 and 14). The amount of vorticity within
the two distinct lobes convected along the cylinder walls increases with β. Indeed,
at β = 400 there is still some vorticity remaining at phase t0 + 0.25T after flow
reversal has started and the counter-rotating vortex pair at the right end is convected
Downloaded from http:/www.cambridge.org/core. University of Georgia Libraries, on 03 Jan 2017 at 13:40:49, subject to the Cambridge Core terms of use, available at
http:/www.cambridge.org/core/terms. http://dx.doi.org/10.1017/jfm.2016.319
 Oscillatory flow past elliptic cylinders 389

(a) 1.0 (b) 1.0

0.5 0.5
0 0
–0.5 –0.5
–1.0 –1.0
–2 –1 0 1 2 –2 –1 0 1 2

(c) 1.0 (d) 1.0

0.5 0.5
0 0
–0.5 –0.5
–1.0 –1.0
–2 –1 0 1 2 –2 –1 0 1 2

(e) 1.0
0.5
0
–0.5
–1.0
–2 –1 0 1 2

F IGURE 14. (Colour online) As figure 11 but β = 400.

in the negative x-direction. These structures also keep the same relative locations
while moving left and right as the flow direction changes. At phase t0 + 0.125T the
free-stream velocity is close to zero and the acceleration of the flow is a maximum,
resulting in expanded vortex cores. Although the structure of the flow at β = 300 and
400 resembles remarkably that at β = 200 in figure 11, the structure of the flow is
now three-dimensional due to the instability in the boundary-layer flow (figure 7a,b).
This similarity in the distribution of the ωz -contours indicates that two-dimensional
flow dominates; moreover the flow retains the symmetry about the (x, z)-plane. The
phase at which the structures are largest (t0 + 0.125T) coincides with the occurrence
of the three-dimensional instability (figure 9). It is also clear from figures 13 and 14
that no vortex shedding occurs.
Next we proceed to explore the properties of the cross-sectional flow at β = 500
and 600 in figures 15 and 16, respectively. At higher β the Stokes layers become
thinner in accordance with the estimate (5.2) while preserving the main flow structure
with counter-rotating vortex pairs generated each half-cycle. The symmetry about
the (x, z)-plane is now broken, which is expected from the spanwise variations of
the flow previously seen in figure 7(c,d). At β = 500 in figure 15, the two distinct
lobes convected along the cylinder walls shrink in size and concentrate more vorticity
than at lower β. Despite this increase in vortex strength the lobes are annihilated at
phase t0 + 0.25T due to strong opposite vorticity travelling in the opposite direction.
The situation is different for β = 600 in figure 16, where detached vortex cores give
evidence of separated flow from the cylinder walls; this is clear from the plots at
phases t0 and t0 + 0.125T. An interesting interaction between the weakened vortex
travelling to the right and the stronger vortex travelling to the left is seen starting
Downloaded from http:/www.cambridge.org/core. University of Georgia Libraries, on 03 Jan 2017 at 13:40:49, subject to the Cambridge Core terms of use, available at
http:/www.cambridge.org/core/terms. http://dx.doi.org/10.1017/jfm.2016.319
 390 J. P. Gallardo, H. I. Andersson and B. Pettersen

(a) 1.0 (b) 1.0

0.5 0.5
0 0
–0.5 –0.5
–1.0 –1.0
–2 –1 0 1 2 –2 –1 0 1 2

(c) 1.0 (d ) 1.0

0.5 0.5
0 0
–0.5 –0.5
–1.0 –1.0
–2 –1 0 1 2 –2 –1 0 1 2

(e) 1.0
0.5
0
–0.5
–1.0
–2 –1 0 1 2

F IGURE 15. (Colour online) As figure 11 but β = 500.

from phase t0 + 0.125T in the upper half of the ellipse. At phase t0 + 0.25T the
two vortices collide at about x/b = 1, and later at phase t0 + 0.375T we observe at
φ = π/2 that the two vorticity regions have merged, resulting in filamentation of the
region with positive vorticity. Detached vorticity regions at both tips of the elliptic
cylinder can also be observed in figure 16.
In order to explore the development of the three-dimensional instability, we use
vorticity contours at vertical planes in the vicinity of the location where the three-
dimensional instability arises. We start by looking at stable structures in the phase
II for β = 300 and 400 indicated in figure 7(a,b), respectively. Previous observations
of equivalent structures on circular cylinders have shown distinctive mushroom-like
shapes corresponding to vortex pairs (see e.g. Sarpkaya 2002; An et al. 2011; Suthon
& Dalton 2011) which are referred to as Honji vortices.
Figures 17 and 18 show the projected normalized vorticity ωn̂ over a cross-section
located at the critical angle φ2 for β = 300 and 400, respectively. The evolution of the
three-dimensional structures is followed through one cycle because, contrary to the
Honji vortices, the structures at the elliptic cylinder with Λ = 1/3 are not sustained
throughout one cycle. In accordance with Hall’s theory, during the first half-cycle
the three-dimensional structures grow and develop at the location φ2 , and the same
process repeats itself during the second half-cycle at φ3 . At phase t0 we observe a
pair of counter-rotating vortices and regions of relatively weak vorticity next to the
wall. Phase t0 + 0.125T corresponds to the region of maximum flow acceleration
and enlargement of the vortex cores in the cross-sectional planes, and here the flow
structures in figures 17 and 18 clearly concentrate more vorticity. The position of the
vortex pair is shifted away from the wall, and a unique pattern consisting of eight
Downloaded from http:/www.cambridge.org/core. University of Georgia Libraries, on 03 Jan 2017 at 13:40:49, subject to the Cambridge Core terms of use, available at
http:/www.cambridge.org/core/terms. http://dx.doi.org/10.1017/jfm.2016.319
 Oscillatory flow past elliptic cylinders 391

(a) 1.0 (b) 1.0

0.5 0.5
0 0
–0.5 –0.5
–1.0 –1.0
–2 –1 0 1 2 –2 –1 0 1 2

(c) 1.0 (d ) 1.0

0.5 0.5
0 0
–0.5 –0.5
–1.0 –1.0
–2 –1 0 1 2 –2 –1 0 1 2

(e) 1.0
0.5
0
–0.5
–1.0
–2 –1 0 1 2

F IGURE 16. (Colour online) As figure 11 but β = 600.

cells with opposite vorticity sign is observed below the vortex pair. Then at phase
t0 + 0.25T the main vortex pair approaches the wall and intense vorticity regions are
seen next to the wall; at this stage we have flow reversal, and the development of the
structures on the opposite side of the ellipse, at φ3 , begins. Thereafter the free-stream
velocity reaches its maximum magnitude at phase t0 + 0.375T, and the vortex pair
gradually decays in the following phases until the growth of the vortex pairs starts
again and the whole process repeats itself. Dramatic growth of the instability will
start again from phase t0 + 0.875T, which is the stage at which the free-stream
velocity reaches its maximum magnitude. In the same way as for a circular cylinder,
it is clearly the vorticity in the boundary layer that provides the mechanical energy
required to sustain the upper counter-rotating vortex pair.
Figures 17 and 18 show the same structure of the evolution of the three-dimensional
instability at β = 300 and 400, the only difference being the increase in the magnitude
of the vorticity at β = 400. In order to quantify this the vorticity has been integrated
over the right half of the planes in figures 17 and 18 to obtain the circulation Γ ; this
is also an alternative way to portray the evolution in time of the three-dimensional
instability. Figure 19 shows that the magnitude of the circulation for β = 400 is about
30 % larger than that at β = 300 from phase t0 to t0 + T/4. Then the circulation
decays rapidly, reaching almost zero values at t0 + T/2 and t0 + 5T/8 for both β = 300
and 400. Then the three-dimensional instability starts growing again, with increasing
differences in circulation between β = 300 and 400 as the flow reverses to the positive
x-direction.
The properties of the flow structures when β increases to 500 and 600 are explored
in figures 20 and 21, respectively. In these flow regimes the structures are expected
Downloaded from http:/www.cambridge.org/core. University of Georgia Libraries, on 03 Jan 2017 at 13:40:49, subject to the Cambridge Core terms of use, available at
http:/www.cambridge.org/core/terms. http://dx.doi.org/10.1017/jfm.2016.319
 392 J. P. Gallardo, H. I. Andersson and B. Pettersen

(a) 1.0 (b) 1.0 (c) 1.0 (d ) 1.0 0

0 0
0.8 0.8 0.8 0.8
0.6 0.6 0.6 0 0 0.6
0.4 0.4 0.4 0
0.4
0
0.2 0.2 0.2 0.2
0 0 0
0 0.4 0.8 0 0.4 0.8 0 0.4 0.8 0 0.4 0.8
(e) 1.0 ( f ) 1.0 0 (g) 1.0 (h) 1.0 0
0.8 0.8 0.8 0.8
0
0.6 0 0.6 0.6 0.6
0
0.4 0.4 0.4 0.4
0
0.2 0.2 0.2 0.2
0 0 0 0
0 0.4 0.8 0 0.4 0.8 0 0.4 0.8 0 0.4 0.8

F IGURE 17. Projected normalized vorticity ωn̂ over a cross-section located at the critical
angle φ2 ; eight different phases at intervals of 1T = 0.125T during a period T. In all the
plots |ωn̂ b/Um | = 2 with 1ωn̂ b/Um = 0.1. Black contours denote positive vorticity and grey
contours negative vorticity. Case with Λ = 1/3, KC = 2 and β = 300.

(a) 1.0 (b) 1.0 (c) 1.0 (d) 1.0

0.8 0.8 0.8 0.8 0
0 0
0 0
0.6 0.6 0.6 0 0.6
0
0.4 0.4 0.4 0.4 0
0 0 0
0.2 0.2 0.2 0
0.2
0 0 0

0 0.4 0.8 0 0.4 0.8 0 0.4 0.8 0 0.4 0.8

0
(e) 1.0 ( f ) 1.0 (g) 1.0 (h) 1.0 0
0.8 0.8 0.8 0.8 0
0
0.6 0.6 0.6 0.6
0 0
0.4 0.4 0.4 0.4
0 0 0
0 0
0.2 0.2 0.2 0.2
0 0 0 0

0 0.4 0.8 0 0.4 0.8 0 0.4 0.8 0 0.4 0.8

F IGURE 18. As figure 17 but β = 400.

to be unstable according to the information conveyed by figure 7(c,d). Several

events where the strips merge can be observed, indicating coalescence of two flow
structures. Using high-speed imagers and laser-induced fluorescence, Sarpkaya (2002)
observed such merging events in the sinusoidal oscillatory flow over a circular
cylinder. Subsequently An et al. (2011) examined the evolution of the interaction
between the mushroom-shaped structures by means of direct numerical simulations.
In figures 20 and 21 the vorticity ωn̂ normal to a plane oriented at φ = 22.5◦ has
been plotted over five periods all at the same phase T/4. The sampling windows are
indicated by the boxes in figure 7(c,d) for β = 500 and 600, respectively. At t = 35.25T
in figure 20 we observe five pairs of counter-rotating vortices along a spanwise length
of 4b. The structure located at midspan (z/b = 0) remains locked at this position as
time advances, exhibiting only minor alterations in its shape. In the left part of the
plots we observe two structures P1 and P2 at t = 35.25T approaching each other.
Positive vorticity in P1 and negative vorticity in P2 cancel each other out during this
Downloaded from http:/www.cambridge.org/core. University of Georgia Libraries, on 03 Jan 2017 at 13:40:49, subject to the Cambridge Core terms of use, available at
http:/www.cambridge.org/core/terms. http://dx.doi.org/10.1017/jfm.2016.319
 Oscillatory flow past elliptic cylinders 393

0.10

0.08

0.06

0.04

0.02

–0.02

F IGURE 19. Integrated normalized circulation Γ over a cross-section located at the critical
angle φ2 . The integration is done from z/b = 0.5 to 1.0 in figures 17 and 18. Cases with
Λ = 1/3, KC = 2, and β = 300 and β = 400.

(a) (b)
0.8 0.8
0.6 P1 P2 P3 P4 0.6
0.4 0.4
0.2 0.2
–2.0 –1.5 –1.0 –0.5 0 0.5 1.0 1.5 2.0 –2.0 –1.5 –1.0 –0.5 0 0.5 1.0 1.5 2.0

(c) (d )
0.8 0.8
0.6 0.6 P5 P3 P4
0.4 0.4
0.2 0.2
–2.0 –1.5 –1.0 –0.5 0 0.5 1.0 1.5 2.0 –2.0 –1.5 –1.0 –0.5 0 0.5 1.0 1.5 2.0

(e)
0.8 P5 P6
0.6
0.4
0.2
–2.0 –1.5 –1.0 –0.5 0 0.5 1.0 1.5 2.0

F IGURE 20. Time evolution of the projected normalized vorticity ωn̂ over a cross-section
located at φ = 22.5◦ (≈6◦ anticlockwise from the critical angle φ2 ). Consecutive snapshots
taken at phase T/4. In all the plots |ωn̂ b/Um | = 2 with 1ωn̂ b/Um = 0.1. Case with Λ = 1/3,
KC = 2 and β = 500.

merging process, and this blending results in a new vortex pair P5 at t = 38.25T. A
similar situation is observed to the right part of the panels, where the pairs P3 and
P4 have merged at t = 39.25T, giving rise to the new vortex pair P6.
Vortex pairs are still seen at β = 600, however the irregularity of their shape
increases considerably (figure 21). Furthermore interactions between the flow
structures will tend to happen more often as seen in figure 7(d). Figure 21 shows
four vortex pairs, Q1–Q4 at 30.25T. The flow structures Q2 and Q3 approach each
other and amalgamate into a new vortex pair Q5 at 32.25T. On the left side the pair
Q1 remains relatively stable throughout the five cycles. The vortex pair Q4 on the
right side exhibits a distorted shape with detached vorticity regions such as that seen
at 32.25T; the shape of this structure is far from being symmetric in contrast to that
observed at lower β. An et al. (2011) also discussed the emergence of new vortices
Downloaded from http:/www.cambridge.org/core. University of Georgia Libraries, on 03 Jan 2017 at 13:40:49, subject to the Cambridge Core terms of use, available at
http:/www.cambridge.org/core/terms. http://dx.doi.org/10.1017/jfm.2016.319
 394 J. P. Gallardo, H. I. Andersson and B. Pettersen

(a) (b)
0.8 0.8
0.6 Q1 Q2 Q3 Q4 0.6
0.4 0.4
0.2 0.2
–2.0 –1.5 –1.0 –0.5 0 0.5 1.0 1.5 2.0 –2.0 –1.5 –1.0 –0.5 0 0.5 1.0 1.5 2.0
(c) (d)
0.8 0.8
0.6 Q1 Q4 0.6
0.4 Q5 0.4
0.2 0.2
–2.0 –1.5 –1.0 –0.5 0 0.5 1.0 1.5 2.0 –2.0 –1.5 –1.0 –0.5 0 0.5 1.0 1.5 2.0

(e)
0.8
0.6 Q1 Q5 Q4
0.4 Q6
0.2
–2.0 –1.5 –1.0 –0.5 0 0.5 1.0 1.5 2.0

F IGURE 21. As figure 20 but β = 600.

when the spacing between two vortices is substantially larger than the average spacing
between contiguous vortices. This is observed at 34.25T between x/b = 0.5 and 1.0
in figure 21, where an incipient flow structure labelled as Q6 can be observed. By
further increasing the frequency parameter β the originally well-organized structure
of vortices will eventually break down.

6. Concluding remarks
In this work we have investigated the oscillating flow past elliptic cylinders by
directly integrating the three-dimensional Navier–Stokes equations forward in time.
Our main objective has been to explore flow regimes corresponding to the early
development of three-dimensional instabilities, and to compare our results with a
previous instability theory for oscillating flow around elliptic cylinders (Hall 1984).
Then, a more in-depth analysis of the flow configuration with highest ellipticity
(Λ = 1/3) has been conducted for a range of frequency parameters β in order to
understand the processes of vortex formation.
For the high-eccentricity case Λ = 1/3 the flow field remained two-dimensional
at the lowest Stokes number β = 200 whereas the originally two-dimensional flow
evolved into a three-dimensional flow field for the higher Stokes numbers in table 1.
Moreover, the higher the β-values considered, the earlier the departure from strict
two-dimensionality and the more irregular the resulting flow field observed, as shown
in figure 7. It should be noted, however, that the ratio between the global Taylor
number Ta and the critical value of the global Taylor number for a circular cylinder
Ta0 varied from 11.7 for β = 200 to 20.3 for β = 600. This implies that the simulated
flow field became unstable at a Ta/Ta0 -value about 10 times higher than the stability
limit that resulted from Hall’s theoretical considerations and reproduced here as the
lower branch in figure 1(a). One should recall that Hall’s linear stability analysis
identified the conditions for marginal stability at one specific location and at the
particular instant of time that corresponded to maximum flow velocity. It is likely,
however, that for a three-dimensional instability to grow, the unstable conditions
Downloaded from http:/www.cambridge.org/core. University of Georgia Libraries, on 03 Jan 2017 at 13:40:49, subject to the Cambridge Core terms of use, available at
http:/www.cambridge.org/core/terms. http://dx.doi.org/10.1017/jfm.2016.319
 Oscillatory flow past elliptic cylinders 395
have to extend over a finite region in space and over a finite time interval. If
so, a significantly higher Taylor number than that predicted by the theory will
be required for a minute disturbance of the two-dimensional flow field to evolve
in time.
Visualizations of vorticity isosurfaces
√ taken at the same phase for five different
aspect ratios of Λ = 1/3, 1/2, 2/3, 3/5 and 1 have shown that the main flow
structure originally identified as ‘streaked flow’ by Honji (1981) is retained for
elliptic
√ cylinders. As expected, the configurations with low eccentricity (Λ = 2/3 and
3/5) exhibited rib-like structures in staggered bands at each side of the cylinder,
similar to those previously observed for circular cylinders (see e.g. Honji 1981;
Tatsuno & Bearman 1990; An et al. 2011; Suthon & Dalton 2012). When the
aspect ratio was decreased further to Λ = 1/2 the vortex pairs experienced a shifting
towards the stagnation point, in accordance with the theoretical predictions (Hall
1984). For Λ = 1/3 the vortical structures were distorted into an oblate shape with
an antisymmetric distribution with respect to the (x, z)-plane. An interesting feature
for Λ = 1/3 and 1/2 was the presence of a pair of counter-rotating spanwise vortex
cores that originates near the tip of the ellipse due to increased local curvature for
low Λ.
The spatio-temporal evolution of the flow structures was examined in detail for
Λ = 1/3 with fixed KC = 2.0 and Stokes number in the range 200 6 β 6 600. The
flow at β = 200 was characterized by the absence of three-dimensional structures; it is
worth mentioning that at this (KC, β)-combination a circular cylinder already exhibits
Honji vortices. The basic two-dimensional structure for this case was tracked during
a half-cycle, revealing Stokes layers with opposite vorticity as previously reported for
oscillating flow around circular cylinders. Indeed, the agreement with Stokes’ solution
for the flow over an oscillating flat plate (Stokes’ second problem) is very good up
to 0.05b from the cylinder wall. We also observed two counter-rotating vortex cores
which form near the tip of the ellipse and are convected along the cylinder walls by
the ambient flow. These two regions of concentrated vorticity gradually diffuse as they
approach the opposite tip of the ellipse, but their strength increases with β.
The basic two-dimensional flow experiences three-dimensional instabilities when
the frequency parameter β is increased further to 300 and 400. The flow structures
are stable and evenly distributed along the cylinder span, resembling the arrangement
observed in oscillatory flows around circular cylinders. Contrary to the circular
cylinder case, however, the evolution of the three-dimensional flow structures during
one cycle of oscillation shows an initial growth up to ≈T/4 followed by decay of
circulation as the instability develops near the opposite tip of the ellipse. The four
locations at which the three-dimensional instability starts to develop for Λ = 1/3 are
in remarkably good agreement with Hall’s theory. The case with β = 400 experiences
a drifting of the flow structures along the span after sufficiently long simulation time.
With a further increase of β to 500 and 600 the structures exhibit merging of regions
with opposite vorticity as well as generation of new pairs of vortices.
It is finally noteworthy that a complete period T was required to portray the time
evolution of the three-dimensional flow structures for an elliptic cylinder, as shown in
figures 17 and 18, whereas only a half-period T/2 is needed for a circular cylinder;
see e.g. figure 13 in An et al. (2011). Moreover, while the Honji vortices that occurred
on a circular cylinder persisted throughout the entire oscillation period (see figure 12
in An et al. 2011), the circulation associated with the vortical structures on the elliptic
cylinder was not sustained throughout the period (see figure 19).
Downloaded from http:/www.cambridge.org/core. University of Georgia Libraries, on 03 Jan 2017 at 13:40:49, subject to the Cambridge Core terms of use, available at
http:/www.cambridge.org/core/terms. http://dx.doi.org/10.1017/jfm.2016.319
 396 J. P. Gallardo, H. I. Andersson and B. Pettersen
Acknowledgements
The authors are grateful to the anonymous referees for their significant and
constructive comments. This work has received support from the Research Council
of Norway (Program for Supercomputing) through a grant of computing time.

REFERENCES

A N , H., C HENG , L. & Z HAO , M. 2011 Direct numerical simulation of oscillatory flow around a
circular cylinder at low Keulegan–Carpenter number. J. Fluid Mech. 666, 77–103.
BADR , H. M. & K OCABIYIK , S. 1997 Symmetrically oscillating viscous flow over an elliptic cylinder.
J. Fluids Struct. 11, 745–766.
B EARMAN , P. W., D OWNIE , M. J., G RAHAM , J. M. R. & O BASAJU , E. D. 1985 Forces on
cylinders in viscous oscillatory flow at low Keulegan–Carpenter numbers. J. Fluid Mech. 154,
337–356.
E LSTON , J. R., B LACKBURN , H. M. & S HERIDAN , J. 2006 The primary and secondary instabilities
of flow generated by an oscillating circular cylinder. J. Fluid Mech. 550, 359–389.
H ALL , P. 1984 On the stability of the unsteady boundary layer on a cylinder oscillating transversely
in a viscous fluid. J. Fluid Mech. 146, 347–367.
H ONJI , H. 1981 Streaked flow around an oscillating circular cylinder. J. Fluid Mech. 107, 509–520.
J EONG , J. & H USSAIN , F. 1995 On the identification of a vortex. J. Fluid Mech. 285, 69–94.
K EULEGAN , G. H. & C ARPENTER , L. H. 1958 Forces on cylinders and plates in an oscillating
fluid. J. Res. Natl Bur. Stand. 60, 423–440.
M ANHART, M. 2004 A zonal grid algorithm for DNS of turbulent boundary layers. Comput. Fluids
33, 435–461.
M ANHART, M., T REMBLAY, F. & F RIEDRICH , R. 2001 MGLET: a parallel code for efficient DNS and
LES of complex geometries. In Parallel Computational Fluid Dynamics-Trends and Applications,
pp. 449–456. Elsevier.
M ORISON , J. R., O’B RIEN , M. P., J OHNSON , J. W. & S CHAAF , S. A. 1950 The force exerted by
surface waves on piles. Petrol. Trans. AIME 189, 149–157.
P ELLER , N., L E D UC , A., T REMBLAY, F. & M ANHART, M. 2006 High-order stable interpolations
for immersed boundary methods. Intl J. Numer. Meth. Fluids 52, 1175–1193.
S ARPKAYA , T. 1986 Force on a circular cylinder in viscous oscillatory flow at low Keulegan–Carpenter
numbers. J. Fluid Mech. 165, 61–71.
S ARPKAYA , T. 2002 Experiments on the stability of sinusoidal flow over a circular cylinder. J. Fluid
Mech. 457, 157–180.
S ARPKAYA , T. 2005 On the parameter β = Re/KC = D2 /νT. J. Fluids Struct. 21, 435–440.
S ARPKAYA , T. 2006 Structures of separation on a circular cylinder in periodic flow. J. Fluid Mech.
567, 281–297.
S TONE , H. L. 1968 Iterative solution of implicit approximations of multidimensional partial differential
equations. SIAM J. Numer. Anal. 5, 530–558.
S UTHON , P. & DALTON , C. 2011 Streakline visualization of the structures in the near wake of a
circular cylinder in sinusoidally oscillating flow. J. Fluids Struct. 27, 885–902.
S UTHON , P. & DALTON , C. 2012 Observations on the Honji instability. J. Fluids Struct. 32, 27–36.
TATSUNO , M. & B EARMAN , P. W. 1990 A visual study of the flow around an oscillating circular
cylinder at low Keulegan–Carpenter numbers and low Stokes numbers. J. Fluid Mech. 211,
157–182.
W HITE , F. M. 2006 Viscous Fluid Flow, 3rd edn. McGraw-Hill.
W ILLIAMSON , C. H. K. 1985 Sinusoidal flow relative to circular cylinders. J. Fluid Mech. 155,
141–174.
W ILLIAMSON , J. H. 1980 Low-storage Runge–Kutta schemes. J. Comput. Phys. 56, 48–56.

Downloaded from http:/www.cambridge.org/core. University of Georgia Libraries, on 03 Jan 2017 at 13:40:49, subject to the Cambridge Core terms of use, available at
http:/www.cambridge.org/core/terms. http://dx.doi.org/10.1017/jfm.2016.319
 Oscillatory flow past elliptic cylinders 397
YANG , K. 2014 Oscillatory flow past cylinders at low KC numbers. PhD thesis, The University of
Western Australia.
YANG , K., C HENG , L., A N , H., BASSOM , A. P. & Z HAO , M. 2014 Effects of an axial flow component
on the Honji instability. J. Fluids Struct. 49, 614–639.
Z HANG , J. & DALTON , C. 1999 The onset of three-dimensionality in an oscillating flow past a fixed
circular cylinder. Intl J. Numer. Meth. Fluids 42, 19–42.

Downloaded from http:/www.cambridge.org/core. University of Georgia Libraries, on 03 Jan 2017 at 13:40:49, subject to the Cambridge Core terms of use, available at
http:/www.cambridge.org/core/terms. http://dx.doi.org/10.1017/jfm.2016.319