scispace - formally typeset
Search or ask a question
Journal ArticleDOI

Two types of axisymmetric helical magnetorotational instability in rotating flows with positive shear

TL;DR: In this paper, a new type of linear axisymmetric helical magnetorotational instability which is capable of destabilizing viscous and resistive rotational flows with radially increasing angular velocity, or positive shear was revealed and investigated.
Abstract: We reveal and investigate a new type of linear axisymmetric helical magnetorotational instability which is capable of destabilizing viscous and resistive rotational flows with radially increasing angular velocity, or positive shear. This instability is double-diffusive by nature and is different from the more familiar helical magnetorotational instability, operating at positive shear above the Liu limit, in that it works instead for a wide range of the positive shear when ${\rm (i)}$ a combination of axial/poloidal and azimuthal/toroidal magnetic fields is applied and ${\rm (ii)}$ the magnetic Prandtl number is not too close to unity. We study this instability first with radially local WKB analysis and then confirm its existence using a global stability analysis of the magnetized flow between two rotating cylinders with conducting or insulating boundaries. From an experimental point of view, we also demonstrate the presence of the new instability in a magnetized viscous and resistive Taylor-Couette flow with positive shear for such values of the flow parameters, which can be realized in upcoming experiments at the DRESDYN facility. Finally, this instability might have implications for the dynamics of the equatorial parts of the solar tachocline and dynamo action there, since the above two necessary conditions for the instability to take place are satisfied in this region. Our global stability calculations for the tachocline-like configuration, representing a thin rotating cylindrical layer with the appropriate boundary conditions -- conducting inner and insulating outer cylinders -- and the values of the flow parameters, indicate that it can indeed arise in this case with a characteristic growth time comparable to the solar cycle period.

Summary (3 min read)

Introduction

  • This is a repository copy of Two types of axisymmetric helical magnetorotational instability in rotating flows with positive shear.
  • Version: Accepted Version Article: Mamatsashvili, G, Stefani, F, Hollerbach, R et al. (1 more author) (2019) Two types of axisymmetric helical magnetorotational instability in rotating flows with positive shear.
  • These conditions are indeed satisfied in the solar tachocline, where this new instability can possibly play an important role in its dynamics and magnetic activity.

II. MAIN EQUATIONS

  • Consider a flow between two coaxial cylinders at inner, ri, and outer, ro, radii, rotating, respectively, with angular velocities Ωi and Ωo in the cylindrical coordinates (r, φ, z).
  • The pressure associated with this base flow and maintaining its rotation is denoted as P0.
  • The imposed background helical magnetic field B0 = (0, B0φ(r), B0z) consists of a radially varying, current-free azimuthal component, B0φ(r) = βB0zro/r, and a constant axial component, B0z , where the constant parameter β characterizes field’s helicity.
  • There is instability in the flow, if the real part of any eigenvalue, or growth rate is positive, Re(γ) > 0, for any of the eigennmodes.

III. WKB ANALYSIS

  • The authors use a radially local WKB approximation, where the radial dependence of the perturbations is assumed to be of the form ∝ exp(ikrr) with kr being the radial wavenumber.
  • Figure 1(a) shows the growth rate, Re(γ), as a function of the re-scaled axial wavenumber, as determined from a numerical solution of Eq. (4) at finite but very small Pm = 10−6, together with solution (5) in the inductionless limit, for fixed Ha∗ and Re∗.
  • Energy is drawn from the background flow rΩ(r) to the growing perturbations due to the coupling between meridional circulation and azimuthal field perturbations brought about by the imposed azimuthal field, a mechanism also underlying HMRI at negative shear [6, 29].
  • To explore the behavior of type 2 Super-HMRI further, the authors first vary α as well as the re-scaled Hartmann and Reynolds numbers.
  • Note that while the authors have only presented the Ro = 1.5 case in order to demonstrate the behavior of the instability with Pm, other values of Rossby number yield qualitatively similar behavior and scalings of γm, Ha ∗ m, Re ∗ m.

IV. GLOBAL ANALYSIS

  • After the radially local WKB analysis, the authors now investigate type 2 Super-HMRI in the global case.
  • Hence, the main equations are the same as those in that paper and are obtained by linearizing Eqs. (1)-(3) about the above equilibrium, with the only difference being that now the imposed rotation profile Ω(r) increases with radius corresponding to outer cylinder rotating faster than the inner one (super-rotation).
  • The boundary conditions are no-slip for the velocity and either perfectly conducting or insulating for the magnetic field.
  • The global 1D analysis serves two main purposes.
  • First, to compare with the results of the above WKB analysis in the regime where the latter holds, i.e., at high β ≫ 10 and a small gap width between the cylinders, δ ≡ ro − ri ≪ ro, when the equilibrium quantities do not change much with radius across the gap.

A. Narrow-gap case

  • One of the main difficulties when comparing local and global analysis of HMRI and AMRI in a magnetic TC flow is that the local Rossby number, which defines these instabilities and to which they are therefore sensitive [19, 23, 24], varies with radius, even in the narrow gap case.
  • This implies that actually the WKB regime is better fulfilled at large Pm and those higher kz , at which the instability reaches a maximum growth, and hence specific boundary conditions do not affect its growth and frequency.
  • This indicates that type 2 Super-HMRI, like HMRI at negative , is a genuine instability intrinsic to the flow, tapping free energy of differential rotation , and is not induced/driven by the radial boundaries on the confining rotating cylinders, although it can still be modified by these boundaries.

B. Wide-gap case

  • Having explored type 2 Super-HMRI in the local and narrow-gap global cases, the authors now look for it in the case of a wide gap.
  • As seen in Fig. 4, qualitatively the expected scalings with β are still followed, with Ha and Re (or S and Rm) as well as kzm increasing with β.
  • By contrast, the local WKB analysis better applies in the narrow gap case, η̂ → 1, because the eigenfunctions vary mostly only over the gap width (Figs. 8-10), which is much less than ro and hence the radial variation of the equilibrium quantities are small across this distance.
  • A more comprehensive global linear analysis of this instability, exploring the dependence of its growth rate on the flow parameters (η̂, β, Re,Ha, Pm), will be presented elsewhere.

V. SUMMARY AND DISCUSSION

  • The authors have uncovered and analyzed a novel type of double-diffusive axisymmetric HMRI, labelled type 2 Super-HMRI, which exists in rotational flows with radially increasing angular velocity, or positive shear of arbitrary steepness threaded by a helical magnetic field – a configuration where magnetorotational instabilities were previously unknown.
  • This will also allow us to make comparison with theoretical results.

A. Applicability to the solar tachocline

  • MRI has already been discussed in relation to the solar tachocline in several studies [17, 36–39], showing that it can arise at middle and high latitudes, where the shear of the differential rotation is negative.
  • This new double-diffusive type 2 Super-HMRI, on the other hand, may arise and potentially have important implications for the dynamics and magnetic activity of the low latitude, near-equatorial region of the solar tachocline, since necessary conditions for the development of this instability can be realized there.
  • It seen from this figure that the largest growth rate is ∼ 10−3, or in dimensional variables ∼ 10−3Ωo, and therefore the corresponding growth time is ∼ 103Ω−1o .

Acknowledgments

  • This project has received funding from the European Union’s Horizon 2020 research and innovation pro- gramme under the Marie Sk lodowska – Curie Grant Agreement No. 795158 and the ERC Advanced Grant Agreement No. 787544 as well as from the Shota Rustaveli National Science Foundation of Georgia (SRNSFG, grant No. FR17-107).
  • GM acknowledges support from the Alexander von Humboldt Foundation .
  • The authors thank both anonymous referees for constructive criticism which has led to an extended version of this paper and much improved the presentation of their main results. [1].
  • Actually the selected value of kr0 when comparing local WKB and global dispersion curves is still somewhat arbitrary, since it is not generally possible to get an exact matching between the characteristic radial length of the global eigenfunctions and radial wavenumber of WKB solutions/harmonics.

Did you find this useful? Give us your feedback

Figures (12)

Content maybe subject to copyright    Report

This is a repository copy of Two types of axisymmetric helical magnetorotational instability
in rotating flows with positive shear.
White Rose Research Online URL for this paper:
http://eprints.whiterose.ac.uk/151156/
Version: Accepted Version
Article:
Mamatsashvili, G, Stefani, F, Hollerbach, R et al. (1 more author) (2019) Two types of
axisymmetric helical magnetorotational instability in rotating flows with positive shear.
Physical Review Fluids, 4 (10). 103905. ISSN 2469-990X
https://doi.org/10.1103/PhysRevFluids.4.103905
© 2019 American Physical Society. This is an author produced version of a paper
published in Physical Review Fluids. Uploaded in accordance with the publisher's
self-archiving policy.
eprints@whiterose.ac.uk
https://eprints.whiterose.ac.uk/
Reuse
Items deposited in White Rose Research Online are protected by copyright, with all rights reserved unless
indicated otherwise. They may be downloaded and/or printed for private study, or other acts as permitted by
national copyright laws. The publisher or other rights holders may allow further reproduction and re-use of
the full text version. This is indicated by the licence information on the White Rose Research Online record
for the item.
Takedown
If you consider content in White Rose Research Online to be in breach of UK law, please notify us by
emailing eprints@whiterose.ac.uk including the URL of the record and the reason for the withdrawal request.

arXiv:1810.13433v2 [physics.flu-dyn] 19 Sep 2019
New type of axisymmetric helical magnetorotational instability in rotating flows with
positive shear
George Mamatsashvili
Niels Bohr International Academy, Niels Bohr Institute,
Blegdamsvej 17, 2100 Copenhagen, Denmark and
Helmholtz-Zentrum Dresden-Rossendorf, Bautzner Landstr. 400, D-01328 Dresden, Germany
Frank Stefani
Helmholtz-Zentrum Dresden-Rossendorf, Bautzner Landstr. 400, D-01328 Dresden, Germany
Rainer Hollerbach
Department of Applied M athematics, University of Leeds, Leeds LS2 9JT, U.K.
unther udiger
Leibniz-Institut ur Astrophysik Potsdam, An der Sternwarte 16, D-14482 Potsdam, Germany
(Dated: September 20, 2019)
We reveal and investigate a new type of linear axisymmetric helical magnetorotational instability
which is capable of destabilizing viscous and resistive rotational flows with radially increasing angular
velocity, or positive shear. This instability is double-diffusive by nature and is different from the
more familiar helical magnetorotational instability, operating at positive shear above the Liu limit, in
that it works instead for a wide range of the positive shear when (i) a combination of axial/poloidal
and azimuthal/toroidal magnetic fields is applied and (ii) the magnetic Prandtl number is not
too close to unity. We study this instability first with radially local WKB analysis, deriving the
scaling properties of its growth rate with respect to Hartmann, Reynolds and magnetic Prandtl
numbers. Then we confirm its existence using a global stability analysis of the magnetized flow
confined between two rotating coaxial cylinders with purely conducting or insulating boundaries
and compare the results with those of the local analysis. From an experimental point of view, we
also demonstrate the presence of the new instability in a magnetized viscous and resistive Taylor-
Couette flow with positive shear for such values of the flow parameters, which can be realized in
upcoming experiments at the DRESDYN facility. Finally, this instability might have implications
for the dynamics of th e equatorial parts of the solar tachocline and dynamo action there, since
the above two necessary conditions for the instability to take place are satisfied in this region.
Our global stability calculations for the tachocline-like configuration, representing a thin rotating
cylindrical layer with the appropriate boundary conditions conducting inner and insulating outer
cylinders and the values of the flow parameters, indicate that it can indeed arise in this case with
a characteristic growth time comparable to the solar cycle period.
I. INTRODUCTION
According to Rayleigh’s criterion [1], rotating flows of
ideal fluids with radially increasing specific angular mo-
mentum are linearly stable. This result has severe astro-
physical consequences, implying hydrodynamic stability
of Keplerian rotation in accretion disks. Nowadays, the
magnetorotational instability (MRI) [24] is considered
to be the most likely destabilizing mechanism for these
disks, driving radially outward transport of angular mo-
mentum and inward accretion of mass.
The standard MRI (SMRI, with a purely ax-
ial/poloidal magnetic field, [24]), as well as the non-
axisymmetric azimuthal MRI (AMRI, with a purely az-
imuthal/toroidal magnetic field, [5]) and the axisymmet-
ric helical MRI (HMRI, with combined axial and az-
imuthal magnetic fields, [6]) have all been extensively
Electronic address: george.mamatsashvili@nbi.ku.dk
studied theoretically (see a recent review [7] and refer-
ences therein). The inductionless forms of AMRI and
HMRI have also been obtained in liquid metal experi-
ments [810], while unambiguous experimental evidence
for inductive SMRI remains elusive, despite promising
first results [11, 12].
In contrast to Keplerian-like rotation with increas-
ing angular momentum but decreasing angular velocity,
much less attention is usually devoted to flows with in-
creasing angular velocity. Until recently, such flows have
been believed to be strongly stable, even under magnetic
fields. However, for very high enough Reynolds num-
bers Re 10
7
, they can yield non-axisymmetric linear
instability [13]. Apart from this hydrodynamic instabil-
ity, there is also a special type of AMRI operating in
flows with much lower Reynolds number but sufficiently
strong positive shear [1416]. This restriction to strong
shear makes, however, this so-called Super-AMRI astro-
physically less significant. One of few positive shear re-
gions is a portion of the solar tachocline extending ±30
about the Sun’s equator. Even there, the shear measured

2
in terms of Rossby number Ro = r(2Ω)
1
d/dr is only
around 0.7 [17, 18], much less than the so-called upper
Liu limit (ULL) Ro
ULL
= 2(1 +
2) 4.83 [19] required
for Super-AMRI. Another astrophysical system in which
positive shear is expected is the boundary layer between
an accretion disk and its host star [20, 21].
Given a general similarity between AMRI and HMRI
and the universal nature of the Liu limits [22, 23], one
might expect a similar result to hold also for Super-
HMRI. However, as we report in this paper, there exists
a new type of axisymmetric HMRI, which we refer to
as type 2 Super-HMRI, that operates in positive shear
flows with arbitrary steepness, whereas the more famil-
iar HMRI operating only at high enough positive shear
above the Liu limit, Ro > Ro
ULL
, is labelled type 1
Super-HMRI. The only requirements are (i) the presence
of both axial and azimuthal magnetic field components
and (ii) magnetic Prandtl number is neither zero (the
inductionless limit) nor too close to unity. These condi-
tions are indeed satisfied in the solar tachocline, where
this new instability can possibly play an important role
in its dynamics and magnetic activity. Although this re-
quires a detailed separate study and is out of scope of
the present paper, we have also done calculations at the
end of this paper, showing the possibility of occurrence of
this instability for the tachocline-like configuration and
parameters, but still remaining in the framework of cylin-
drical flow. The resulting growth time (inverse of the ex-
ponential growth rate) of the most unstable mode in fact
turns out to be comparable to the solar cycle period.
In this paper, we carry out a linear stability analy-
sis of a magnetic rotational flow in cylindrical geome-
try mainly using the Wentzel-Kramers-Brillouin (WKB)
short-wavelength formulation of the underlying magneto-
hydrodynamics (MHD) problem [23, 24], which is espe-
cially useful for understanding the basic features and scal-
ing properties of the new instability. This local analysis
is then complemented by global, radially one-dimensional
(1D) calculations of the corresponding unstable eigen-
modes with the primary aim to demonstrate the exis-
tence of this new version of Super-HMRI beyond the lo-
cal WKB approximation as well as to draw a compari-
son with the results obtained using this approximation.
A more comprehensive global linear analysis exploring
parameter space, and subsequently nonlinear analysis of
this double-diffusive type 2 Super-HMRI at positive shear
will be presented elsewhere.
The paper is organized as follows. Main equations and
the formulation of a problem are given in Sec. II. The
local WKB analysis of the instability is presented in Sec.
III. The global stability analysis of a differentialy rotating
flow between two coaxial cylinders at positive shear both
in the narrow and wide gap cases as well as a comparison
with the results of the local analysis are presented in Sec.
IV. A summary and discussion on the relevance of this
new version of Super-HMRI to the solar tachocline are
given in Sec. V.
II. MAIN EQUATIONS
The motion of an incompressible conducting medium
with constant viscosity ν and ohmic resistivity η is gov-
erned by the equations of non-ideal MHD
U
t
+(U·)U =
1
ρ
P +
B
2
2µ
0
+
(B · )B
µ
0
ρ
+ν
2
U,
(1)
B
t
= × (U × B) + η
2
B, (2)
· U = 0, · B = 0. (3)
where ρ is the constant density, U is the velocity, P is
the thermal pressure, B is the magnetic field and µ
0
is
the magnetic permeability of vacuum.
Consider a flow between two coaxial cylinders at in-
ner, r
i
, and outer, r
o
, radii, rotating, respectively, with
angular velocities
i
and
o
in the cylindrical coordi-
nates (r, φ, z). Since we are primarily interested in the
flow stability in the case of positive shear, or so-called
“super-rotation” [15, 16], the inner cylinder is assumed
to rotate slower than the outer one,
i
<
o
, induc-
ing an azimuthal nonuniform flow U
0
= (0, r(r), 0) be-
tween the cylinders with radially increasing angular ve-
locity, d/dr > 0, and hence positive Rossby number,
Ro > 0. The pressure associated with this base flow and
maintaining its rotation is denoted as P
0
. The imposed
background helical magnetic field B
0
= (0, B
0φ
(r), B
0z
)
consists of a radially varying, current-free azimuthal com-
ponent, B
0φ
(r) = βB
0z
r
o
/r, and a constant axial compo-
nent, B
0z
, where the constant parameter β characterizes
field’s helicity.
We investigate the linear stability of this equilibrium
against small axisymmetric (/∂φ = 0) perturbations,
u = U U
0
, p = P P
0
, b = B B
0
, which are all
functions of r and depend on time t and axial/vertical z-
coordinate via exp(γt+ik
z
z), where γ is the (complex)
eigenvalue and k
z
is the axial wavenumber. There is in-
stability in the flow, if the real part of any eigenvalue, or
growth rate is positive, Re(γ) > 0, for any of the eigen-
nmodes. In such cases, for a given set of parameters, we
always select out the mode with the largest growth rate
from a corresponding eigenvalue spectrum.
III. WKB ANALYSIS
In this section, we use a radially local WKB approx-
imation, where the radial dependence of the perturba-
tions is assumed to be of the form exp(ik
r
r) with k
r
being the radial wavenumber. The resulting dispersion
relation, which follows from Eqs. (1)-(3) after linearizing
and substituting the above exponential form of the per-
turbations, is represented by the fourth-order polynomial
[23, 24]:
γ
4
+ a
1
γ
3
+ a
2
γ
2
+ (a
3
+ ib
3
)γ + a
4
+ ib
4
= 0, (4)

3
with the real coefficients
a
1
= 2
k
2
Re
1 +
1
P m
,
a
2
= 4α
2
(1 + Ro) + 2(k
2
z
+ 2α
2
β
2
)
Ha
2
Re
2
P m
+
k
4
Re
2
1 +
4
P m
+
1
P m
2
,
a
3
= 8(1 + Ro)α
2
k
2
ReP m
+ 2[k
4
+ (k
2
z
+ 2α
2
β
2
)Ha
2
]
k
2
Re
3
P m
1 +
1
P m
b
3
= 8α
2
βk
z
Ha
2
Re
2
P m
,
a
4
= 4α
2
k
4
P m
2
(1 + Ro)
1
Re
2
+ β
2
Ha
2
Re
4
+ 4α
2
k
2
z
Ro
Ha
2
Re
2
P m
+
k
2
z
Ha
2
+ k
4
2
1
Re
4
P m
2
,
b
4
= 4βk
3
z
Ro
1
1
P m
2
P m
Ha
2
Re
3
P m
.
Henceforth γ is normalized by the outer cylinder’s angu-
lar velocity
o
, and the wavenumbers by its inverse ra-
dius, r
1
o
. Other nondimensional parameters are: α =
k
z
/k, where k = (k
2
r
+ k
2
z
)
1/2
is the total wavenum-
ber; the Reynolds number Re =
o
r
2
o
, the magnetic
Reynolds number Rm =
o
r
2
o
, and their ratio, the
magnetic Prandtl number P m = ν/η = Rm/Re; the
Hartmann number Ha = B
0z
r
o
/(µ
0
ρνη)
1/2
that mea-
sures the strength of the imposed axial magnetic field.
Another quantity characterizing the field is Lundquist
number S = Ha · P m
1/2
, which, like Rm, does not in-
volve viscosity. Since we focus on positive Rossby num-
bers, Ro > 0, or positive shear, the flow is generally
stable both hydrodynamically, according to Rayleigh’s
criterion (but see Ref. [13]), as well as against SMRI
with a purely axial field (β = 0) [2426].
In the inductionless limit, P m 0, the roots of Eq.
(4) can be found analytically [19, 23, 24, 27, 28]. For
positive and relatively large Ro > Ro
ULL
, one of the
roots always has a positive real part, implying instability
with the growth rate
Re(γ) =
q
2X + 2
p
X
2
+ Y
2
(k
2
z
+ 2α
2
β
2
)
Ha
2
k
2
Re
Re
k
2
, (5)
where
X = α
2
β
2
(α
2
β
2
+ k
2
z
)
Ha
4
Re
2
k
4
α
2
(1 + Ro),
Y = βα
2
k
z
(2 + Ro)
Ha
2
k
2
Re
,
which we call type 1 Super-HMRI. Our main goal though
is to reveal that apart from this type 1 Super-HMRI at
large positive shear, Eq. (4) also yields a completely new
type of dissipation-induced double-diffusive instability at
finite P m, which we call type 2 Super-HMRI.
Regarding the dependence on β parameter in Eq. (4),
it is readily seen that, as long as β 6= 0, it enters the coef-
ficients of these dispersion relations through the re-scaled
wavenumbers, Hartmann, Lundquist and Reynolds num-
bers, k
z
k
z
, k
k, Ha
Ha/β, S
S /β,
Re
Re/β
2
, Rm
Rm/β
2
, in terms of which we
carry out the following WKB analysis. It is easy to check
that β disappears in the polynomial Eq. (4) after sub-
stituting these re-scaled parameters (denoted with aster-
isks) in its coefficients.
Figure 1(a) shows the growth rate, Re(γ), as a func-
tion of the re-scaled axial wavenumber, as determined
from a numerical solution of Eq. (4) at finite but very
small P m = 10
6
, together with solution (5) in the in-
ductionless limit, for fixed Ha
and Re
. For the Rossby
number we take the values lower, Ro = 1.5, 2, and higher,
Ro = 6, than Ro
ULL
. Two distinct instability regimes
are clearly seen in this figure. Type 2 Super-HMRI is
concentrated at small k
z
and exists at finite P m both for
Ro < Ro
ULL
and Ro > Ro
ULL
, i.e., it is insensitive to
the upper Liu limit, but disappears for P m 0 at fixed
Hartmann and Reynolds numbers. By contrast, type 1
Super-HMRI, concentrated at larger k
z
, exists only for
Ro > R o
ULL
, and approaches the inductionless solution
as P m 0. This latter branch is basically an extension
of the more familiar HMRI operating at negative shear,
which in the inductionless limit also satisfies Eq. (5), but
at Ro < Ro
LLL
, where Ro
LLL
= 2(1
2) 0.83 is
the lower Liu limit [19, 24, 27].
At large P m 1, type 1 Super-HMRI disappears and
there remains only type 2 Super-HMRI, as shown in Fig.
2(a). The corresponding dispersion curves as a function
of axial wavenumber have a shape similar to those at
small P m in Fig. 1(a), but now the instability occurs
at order of magnitude larger k
z
and several orders of
magnitude smaller Ha
and Re
at the same values of
Ro adopted in these figures.
Thus, type 2 Super-HMRI represents a new,
dissipation-induced instability mode at positive shear,
which appears to require the presence of both finite vis-
cosity and resistivity. As we will see below though, it
does not operate in the immediate vicinity of P m = 1,
that is, it is double-diffusive in nature, operating for both
small and large P m, but not for P m = O(1). Just as
all previous MRI variants, this one also derives energy

4
solely from the shear, since the imposed magnetic field
is current-free, thereby eliminating current-driven insta-
bilities, such as the Tayler instability. Energy is drawn
from the background flow rΩ(r) to the growing perturba-
tions due to the coupling between meridional circulation
and azimuthal field perturbations brought about by the
imposed azimuthal field, a mechanism also underlying
HMRI at negative shear [6, 29].
Our main goal is to describe the properties of this new
type 2 Super-HMRI. Type 1 Super-HMRI, existing only
for Ro > Ro
ULL
and persisting even in the inductionless
limit P m 0 [19, 23, 24, 28], is also relatively new and
interesting in its own right, but will not be considered
here further.
Like normal HMRI at negative shear, type 2 Super-
HMRI is an overstability, that is, its growth rate comes
with an associated non-zero imaginary part, ω = Im(γ),
which is the frequency of temporal oscillations of the
solution at a given coordinate and, together with ax-
ial wavenumber, defines its propagation speed. Figure 3
shows these frequencies as a function of k
z
, correspond-
ing to the growth rates plotted in Figs. 1(a) and 2(a).
They monotonically increase with k
z
by absolute value,
but are positive at small P m and negative at large P m,
implying opposite propagation directions of the wave pat-
terns at these magnetic Prandtl numbers. Also, ω re-
mains smaller than the frequency of inertial oscillations,
ω
io
= 2α(1 + Ro)
1/2
, and tend to the latter only at small
P m as the solution changes from type 2 to type 1 Super-
HMRI with increasing k
z
and do not change afterwards.
This reflects the fact that type 1 Super-HMRI represents
weakly destabilized inertial oscillations, like the normal
HMRI at negative shear [19].
To explore the behavior of type 2 Super-HMRI further,
we first vary α as well as the re-scaled Hartmann and
Reynolds numbers. The growth rate, maximized over
the last two numbers and k
z
, increases linearly with α
and scales as Ro
1.75
at small P m = 10
6
(Fig. 1(b))
and as Ro
1.18
at large P m = 100 (Fig. 2(b)), while its
dependence on Ha
and Re
, when maximized over k
z
and α, is shown in Fig. 1(c) at P m = 10
6
and in Fig.
2(c) at P m = 100 with Ro = 1.5 < Ro
ULL
(when type 1
Super-HMRI is absent) in both cases. The most unstable
region is quite localized, with the growth rate decreasing
for both small and large Ha
and Re
, implying that this
instability relies on finite viscosity and resistivity, i.e., it
is indeed of double-diffusive type. The overall shape of
the unstable area in (Ha
, Re
)-plane does not change
qualitatively at other P m and Ro; the unstable region
always remains localized and shifts to larger Ha
and
Re
with decreasing P m. In particular, the maximum
growth rate, γ
m
, occurs for (Ha
m
, Re
m
) (700, 9 · 10
4
)
when Pm is small (Fig. 1(c)), but for orders of magnitude
smaller (Ha
m
, Re
m
) = (2.54, 0.23) when P m is large
(Fig. 2(c)) The actual values of the characteristic vertical
wavenumber, Hartmann and Reynolds numbers for type
2 Super-HMRI at different β are obtained by simply mul-
tiplying the values of re-scaled quantities k
z
, Ha
, Re
10
-2
10
-1
10
0
k
z
10
-3
10
-2
10
-1
Re( )
Ro=1.5
Ro=2
Ro=6
Ro=6, Pm=0
0 0.2 0.4 0.6 0.8 1
0
0.01
0.02
m
/Ro
1.75
10
1
10
2
10
3
10
4
10
5
Ha
10
3
10
4
10
5
10
6
10
7
Re
0.01
0.02
0.03
0.04
Type 2 Super-HMRI
Type 1 Super-HMRI
(b)
(a)
(c)
FIG. 1: Panel (a) shows the growth rate Re(γ) v s. k
z
at
fixed Ha
= 90, Re
= 8 · 10
3
, α = 0.71 (i.e., k
r
= k
z
) and
P m = 10
6
for different Ro = 1.5 (blue ), 2 (green), 6 (red).
New type 2 Super-HMRI branch exists at smaller k
z
and finite
P m, for all three Ro values. By contrast, type 1 Super-HMRI
branch at larger k
z
appears only for Ro = 6 > Ro
ULL
from
these th ere values of the Rossby number, but persists also
in the inductionless limit (Eq. 5, dashed-black line). For the
same P m, panel (b ) shows the growth rate of type 2 Super-
HMRI, maximized over a set of the parameters (k
z
, Ha
,
Re
) and normalized by Ro
1.75
, vs. α, while panel (c) shows
the growth rate, maximized over k
z
and α, as a function of
Ha
and Re
at Ro = 1.5 and the same P m = 10
6
.
10
-2
10
-1
10
0
k
z
10
-3
10
-2
10
-1
10
0
Re( )
Ro=1.5
Ro=2
Ro=6
0 0.2 0.4 0.6 0.8 1
0
0.05
0.1
m
/Ro
1.18
10
-4
10
-2
10
0
10
2
10
4
Ha
10
-4
10
-2
10
0
10
2
Re
0.05
0.1
0.15
0.2
(a)
(c)
Type 2 Super-HMRI
(b)
FIG. 2: Same as in Fig. 1, but at Ha
= 5, Re
= 0.1,
α = 0.71 in panel (a) and P m = 100 in all panels. New
type 2 Super-HMRI branch exists at h igher k
z
than those at
small P m, while type 1 Super-HMRI b ranch is absent. In
panel (b), the maximum growth rate now exhibits the scaling
with Rossby number, Ro
1.18
, different from that at small P m.
In panel (c), the maximum growth occurs now at orders of
magnitude smaller Ha
m
and Re
m
than those at small P m in
Fig. 1(c) at the same Ro = 1.5.

Citations
More filters
Journal ArticleDOI
TL;DR: In this article, a solar dynamo model of the Tayler-Spruit type whose Omega effect is conventionally produced by a solar-like differential rotation but whose alpha effect is periodically modulated by planetary tidal forcing is discussed.
Abstract: We discuss a solar dynamo model of Tayler-Spruit type whose Omega-effect is conventionally produced by a solar-like differential rotation but whose alpha-effect is assumed to be periodically modulated by planetary tidal forcing. This resonance-like effect has its rationale in the tendency of the current-driven Tayler instability to undergo intrinsic helicity oscillations which, in turn, can be synchronized by periodic tidal perturbations. Specifically, we focus on the 11.07 years alignment periodicity of the tidally dominant planets Venus, Earth, and Jupiter, whose persistent synchronization with the solar dynamo is briefly touched upon. The typically emerging dynamo modes are dipolar fields, oscillating with a 22.14 years period or pulsating with a 11.07 years period, but also quadrupolar fields with corresponding periodicities. In the absence of any constant part of alpha, we prove the subcritical nature of this Tayler-Spruit type dynamo. The resulting amplitude of the alpha oscillation that is required for dynamo action turns out to lie in the order of 1 m/s, which seems not implausible for the sun. When starting with a more classical, non-periodic part of alpha, even less of the oscillatory alpha part is needed to synchronize the entire dynamo. Typically, the dipole solutions show butterfly diagrams, although their shapes are not convincing yet. Phase coherent transitions between dipoles and quadrupoles, which are reminiscent of the observed behaviour during the Maunder minimum, can be easily triggered by long-term variations of dynamo parameters, but may also occur spontaneously even for fixed parameters. Further interesting features of the model are the typical second intensity peak and the intermittent appearance of reversed helicities in both hemispheres.

43 citations

Journal ArticleDOI
TL;DR: In this paper , a concise review is given of astrophysically motivated experimental and theoretical research on Taylor-Couette flow, and some outstanding questions and near-future prospects are discussed, especially in connection with astrophysics.
Abstract: A concise review is given of astrophysically motivated experimental and theoretical research on Taylor–Couette flow. The flows of interest rotate differentially with the inner cylinder faster than the outer, but are linearly stable against Rayleigh’s inviscid centrifugal instability. At shear Reynolds numbers as large as 106, hydrodynamic flows of this type (quasi-Keplerian) appear to be nonlinearly stable: no turbulence is seen that cannot be attributed to interaction with the axial boundaries, rather than the radial shear itself. Direct numerical simulations agree, although they cannot yet reach such high Reynolds numbers. This result indicates that accretion-disc turbulence is not purely hydrodynamic in origin, at least insofar as it is driven by radial shear. Theory, however, predicts linear magnetohydrodynamic (MHD) instabilities in astrophysical discs: in particular, the standard magnetorotational instability (SMRI). MHD Taylor–Couette experiments aimed at SMRI are challenged by the low magnetic Prandtl numbers of liquid metals. High fluid Reynolds numbers and careful control of the axial boundaries are required. The quest for laboratory SMRI has been rewarded with the discovery of some interesting inductionless cousins of SMRI, and with the recently reported success in demonstrating SMRI itself using conducting axial boundaries. Some outstanding questions and near-future prospects are discussed, especially in connection with astrophysics. This article is part of the theme issue ‘Taylor–Couette and related flows on the centennial of Taylor’s seminal Philosophical Transactions paper (Part 2)’.

3 citations

References
More filters
Journal ArticleDOI
TL;DR: In this article, the authors examined the MHD instabilities arising in the radiation zone of a differentially rotating star, in which a poloidal field of fossil origin is sheared into a toroidal field.
Abstract: Context. We examine the MHD instabilities arising in the radiation zone of a differentially rotating star, in which a poloidal field of fossil origin is sheared into a toroidal field. Aims. We focus on the non-axisymmetric instability that affects the toroidal magnetic field in a rotating star, which was first studied by Pitts and Tayler in the non-dissipative limit. If such an instability were able to mix the stellar material, it could have an impact on the evolution of the star. According to Spruit, it could also drive a dynamo. Methods. We compare the numerical solutions built with the 3-dimensional ASH code with the predictions drawn from an analytical study of the Pitts & Tayler instability. Results. The Pitts & Tayler instability is manifestly present in our simulations, with its conspicuous m = 1 dependence in azimuth. But its analytic treatment used so far is too simplified to be applied to the real stellar situation. Although the instability generated field reaches an energy comparable to that of the mean poloidal field, that field seems unaffected by the instability: it undergoes Ohmic decline, and is neither eroded nor regenerated by the instability. The toroidal field is produced by shearing the poloidal field and it draws its energy from the differential rotation. The small scale motions behave as Alfven waves; they cause negligible eddy-diffusivity and contribute little to the net transport of angular momentum. Conclusions. In our simulations we observe no sign of dynamo action, of either mean field or fluctuation type, up to a magnetic Reynolds number of 10 5 . However the Pitts & Tayler instability is sustained as long as the differential rotation acting on the poloidal field is able to generate a toroidal field of sufficient strength. But in the Sun such a poloidal field of fossil origin is ruled out by the nearly uniform rotation of the deep interior.

257 citations

Journal ArticleDOI
TL;DR: This work presents the first experimental observation of the magnetorotational instability of liquid sodium between differentially rotating spheres, with an imposed coaxial magnetic field, and characterize the observed patterns, dynamics, and torque increases.
Abstract: Differential rotation occurs in conducting flows in accretion disks and planetary cores. In such systems, the magnetorotational instability can arise from coupling Lorentz and centrifugal forces to cause large radial angular momentum fluxes. We present the first experimental observation of the magnetorotational instability. Our system consists of liquid sodium between differentially rotating spheres, with an imposed coaxial magnetic field. We characterize the observed patterns, dynamics, and torque increases, and establish that this instability can occur from a hydrodynamic turbulent background.

227 citations

Journal ArticleDOI
TL;DR: In this article, the possibility of studying MRI in a rotating liquid metal annulus (Couette flow) is explored by local and global stability analysis, and it is shown that MRI can be triggered in a moderately rapidly rotating table-top apparatus, using easy-to-handle metals such as gallium.
Abstract: Although the magnetorotational instability (MRI) has been widely accepted as a powerful accretion mechanism in magnetized accretion discs, it has not been realized in the laboratory. The possibility of studying MRI in a rotating liquid metal annulus (Couette flow) is explored by local and global stability analysis. Stability diagrams are drawn in dimensionless parameters, and also in terms of the angular velocities at the inner and outer cylinders. It is shown that MRI can be triggered in a moderately rapidly rotating table-top apparatus, using easy-to-handle metals such as gallium. Practical issues of this proposed experiment are discussed.

188 citations

Journal ArticleDOI
TL;DR: Results of a Taylor-Couette experiment with the liquid metal alloy GaInSn are presented, showing evidence for the existence of the magnetorotational instability at Reynolds numbers of order 1000 and HartmannNumbers of order 10.
Abstract: A recent Letter [R. Hollerbach and G. R\"udiger, Phys. Rev. Lett. 95, 124501 (2005)] has shown that the threshold for the onset of the magnetorotational instability in a Taylor-Couette flow is dramatically reduced if both axial and azimuthal magnetic fields are imposed. In agreement with this prediction, we present results of a Taylor-Couette experiment with the liquid metal alloy GaInSn, showing evidence for the existence of the magnetorotational instability at Reynolds numbers of order 1000 and Hartmann numbers of order 10.

180 citations

Journal ArticleDOI
TL;DR: This work studies the stability of cylindrical Taylor-Couette flow in the presence of combined axial and azimuthal magnetic fields, and shows that adding an azIMuthal field profoundly alters the previous results for purely axial fields.
Abstract: We study the stability of cylindrical Taylor-Couette flow in the presence of combined axial and azimuthal magnetic fields, and show that adding an azimuthal field profoundly alters the previous results for purely axial fields. For small magnetic Prandtl numbers Pm, the critical Reynolds number ${\mathrm{Re}}_{c}$ for the onset of the magnetorotational instability becomes independent of Pm, whereas for purely axial fields it scales as ${\mathrm{Pm}}^{\ensuremath{-}1}$. For typical liquid metals, ${\mathrm{Re}}_{c}$ is then reduced by several orders of magnitude, enough that this new design should succeed in realizing this instability in the laboratory.

156 citations

Frequently Asked Questions (17)
Q1. What are the contributions in this paper?

This is indicated by the licence information on the White Rose Research Online record for the item. 

This promising result of the global stability analysis can, in turn, be a basis for future efforts aiming at the detection of this instability and thereby providing an experimental evidence for its existence. Building on the findings of this first study, future work will explore in greater detail the parameter space and the effect of boundary conditions on type 2 Super-HMRI, which are more relevant to those present in TC flow experiments. 

Energy is drawn from the background flow rΩ(r) to the growing perturbations due to the coupling between meridional circulation and azimuthal field perturbations brought about by the imposed azimuthal field, a mechanism also underlying HMRI at negative shear [6, 29]. 

The imposed background helical magnetic field B0 = (0, B0φ(r), B0z) consists of a radially varying, current-free azimuthal component, B0φ(r) = βB0zro/r, and a constant axial component, B0z , where the constant parameter β characterizes field’s helicity. 

A basic condition for the local WKB approximation to hold is that the radial wavelength, λr, of the perturbations must be much smaller than the characteristic radial size, ro, of the system over which the equilibrium quantities vary, i.e., λr ≪ ro. 

Since the authors focus on positive Rossby numbers, Ro > 0, or positive shear, the flow is generally stable both hydrodynamically, according to Rayleigh’s criterion (but see Ref. [13]), as well as against SMRI with a purely axial field (β = 0) [24–26]. 

depending on Ha and Re, conducting boundaries can lead either to increase or decrease of the growth rate compared to its WKB value, whereas insulating boundaries alwaysreduce the growth rate about three times compared to those for the conducting ones for fixed Hartmann and Reynolds numbers. 

Since the authors are primarily interested in the flow stability in the case of positive shear, or so-called “super-rotation” [15, 16], the inner cylinder is assumed to rotate slower than the outer one, Ωi < Ωo, inducing an azimuthal nonuniform flow U0 = (0, rΩ(r), 0) between the cylinders with radially increasing angular velocity, dΩ/dr > 0, and hence positive Rossby number, Ro > 0. 

In this section, the authors use a radially local WKB approximation, where the radial dependence of the perturbations is assumed to be of the form ∝ exp(ikrr) with kr being the radial wavenumber. 

Another reason for the differences in the growth rates between the local and the global wide gap cases is that a finite distance between the cylinders also excludes (cuts off) very small radial wavenumbers (α → 1), which correspond to larger growth rates at β ∼ 1 in the WKB analysis (see Fig. 1(b)). 

As noted above, in this subsection, the primary goal of these calculations is to identify this new instability in a TC flow setup commonly employed in lab experiments, which in the present case has a radially increasing angular velocity (positive shear) profile. 

Regarding the dependence on β parameter in Eq. (4), it is readily seen that, as long as β 6= 0, it enters the coefficients of these dispersion relations through the re-scaled wavenumbers, Hartmann, Lundquist and Reynolds numbers, k∗z ≡ kz/β, k∗ ≡ k/β, Ha∗ ≡ Ha/β, S∗ ≡ S/β, Re∗ ≡ Re/β2, Rm∗ ≡ Rm/β2, in terms of which the authors carry out the following WKB analysis. 

type 2 Super-HMRI represents a new, dissipation-induced instability mode at positive shear, which appears to require the presence of both finite viscosity and resistivity. 

This latter branch is basically an extension of the more familiar HMRI operating at negative shear, which in the inductionless limit also satisfies Eq. (5), but at Ro < RoLLL, where RoLLL = 2(1 − √2) ≈ −0.83 is the lower Liu limit [19, 24, 27]. 

The authors take β ∼ 1 in these calculations for the wide-gap, since it is rather costly to achieve large β in experiments due to very high axial currents required. 

In particular, the maximum growth rate, γm, occurs for (Ha ∗ m, Re ∗m) ≈ (700, 9 · 104) when Pm is small (Fig. 1(c)), but for orders of magnitude smaller (Ha∗m, Re ∗m) = (2.54, 0.23) when Pm is large (Fig. 2(c)) 

the authors will see in the global linear analysis below that this mode of instability is in fact not restricted only to large β and can even exist at smaller β ∼ 1, but in this case the WKB approach is questionable and should be applied with caution.