dipolar dynamos in stratified systemspresentations.dynamos2015.de/raynaud_r.pdf · introduction...
Post on 18-Oct-2020
10 Views
Preview:
TRANSCRIPT
Introduction Modelling Axial dipoles Conclusion
Dipolar dynamos in stratified systems
Raphaël Raynaud, Ludovic Petitdemange, Emmanuel Dormy
LRA-LERMAÉCOLE NORMALE SUPÉRIEURE
Stellar and Planetary DynamosGöttingen – May 27, 2015
1 / 21
Introduction Modelling Axial dipoles Conclusion
Table of contents
1 Introduction
2 ModellingGoverning equationsSet upMethods
3 Axial dipoles
4 Conclusion
2 / 21
Introduction Modelling Axial dipoles Conclusion
Different magnetospheric configurations
Axisymmetric (m = 0) and non-axisymmetric (m = 1) configurations
Ridley, V. A., Jovimagnetic secular variation, Doctoral thesis, 2013
3 / 21
Introduction Modelling Axial dipoles Conclusion
Table of contents
1 Introduction
2 ModellingGoverning equationsSet upMethods
3 Axial dipoles
4 Conclusion
4 / 21
Introduction Modelling Axial dipoles Conclusion
Governing equations
How to model the fluid flow ?
The Convective Approximationsretain the essential physics with a minimum complexity
filter out sound waves
The Boussinesq Approximation
∇· (u)= 0
The Boussinesq benchmark(Christensen et al. 2001)
homogeneous mass distribution
g ∝ r
The Anelastic Approximation
∇· (ρa u)= 0
The anelastic benchmark(Jones et al. 2011)
central mass distribution
g ∝ 1/r2
5 / 21
Introduction Modelling Axial dipoles Conclusion
Governing equations
How to model the fluid flow ?
The Convective Approximationsretain the essential physics with a minimum complexity
filter out sound waves
The Boussinesq Approximation
∇· (u)= 0
The Boussinesq benchmark(Christensen et al. 2001)
homogeneous mass distribution
g ∝ r
The Anelastic Approximation
∇· (ρa u)= 0
The anelastic benchmark(Jones et al. 2011)
central mass distribution
g ∝ 1/r2
5 / 21
Introduction Modelling Axial dipoles Conclusion
Governing equations
The anelastic approximation
Polytropic reference state adiabatically stratified
Ta =w(r), ρa =wn, Pa =wn+1 (1)
Resulting system
∇· (ρaV)= 0 (2)
DtV=−∇(Pc
ρa
)−αSScg+Fν/ρa (3)
ρaDtSc +∇·(
Iq
Ta
)= Q
Ta+ Iq ·∇T−1
a (4)
with
Fν viscous forceQ viscous heatingIq heat flux
Braginsky, S. I. & Roberts, P. H., 1995, GAFD, 79, 1
6 / 21
Introduction Modelling Axial dipoles Conclusion
Governing equations
Mean field approach for the heat transfer equation
Solve the equation for average quantitiesf = ⟨f ⟩+ f t (5)
Turbulent fluctuations create an entropy flux term
ISt = ρa⟨StVt ⟩∝∇⟨S⟩ (6)
Strong simplification
the usual molecular term Iq =−k∇T is then neglected in theheat transfer equation
=⇒ temperature is removed from the problem !
Drawbacksargument only valid far from the onset of convection
P2 of thermodynamics not under warranty anymore
7 / 21
Introduction Modelling Axial dipoles Conclusion
Governing equations
MHD equations
Dimensionless system
∂v∂t
+v ·∇v = Pm[− 1
E∇ P ′
wn + PmPr
RaSr2 er − 2
Eez ×v
+Fν+ 1E wn (∇×B)×B
], (7)
∂B∂t
= ∇× (v×B)+∇2B , (8)
∂S∂t
+v ·∇S = w−n−1 PmPr
∇·(wn+1∇S
)+Di
w
[E−1w−n(∇×B)2 +Qν
], (9)
∇· (wnv) = 0 , (10)
∇·B = 0 . (11)
8 / 21
Introduction Modelling Axial dipoles Conclusion
Set up
A simplified model
Set upA perfect gas in a rotating sphericalshell with constant? kinematic viscosity ν? turbulent entropy diffusivity κ? magnetic diffusivity η
Convection is driven by an imposedentropy difference ∆S between theinner and outer spheres
Boundary conditionsfixed entropy
stress-free b. c. for the velocity field
insulating b. c. for the magnetic field9 / 21
Introduction Modelling Axial dipoles Conclusion
Methods
Systematic parameter study (∼ 250 models)
Seven control parameters
Rayleigh number Ra GMd∆Sνκcp
O(106)
magnetic Prandtl number Pm ν/η O(1)Prandtl number Pr ν/κ 1Ekman number E ν/(Ωd2) 10−4
aspect ratio χ ri/ro 0.35polytropic index n 1/(γ−1) 2number of density scale heights N% ln(%i/%o) ≤ 3
The anelastic version of PaRoDy reproduces the anelastic dynamobenchmark (Jones et al. 2011).
10 / 21
Introduction Modelling Axial dipoles Conclusion
Table of contents
1 Introduction
2 ModellingGoverning equationsSet upMethods
3 Axial dipoles
4 Conclusion
11 / 21
Introduction Modelling Axial dipoles Conclusion
Stability domain of the dipolar branch N% = 0.1
0 5 10 15 20 25 30 35
Ra/Rac
0
1
2
3
4
5
6
7
Pm
Dipolar dynamos as a function of Ra/Rac and Pm12 / 21
Introduction Modelling Axial dipoles Conclusion
Stability domain of the dipolar branch N% = 0.5
0 5 10 15 20 25 30 35
Ra/Rac
0
1
2
3
4
5
6
7
Pm
Dipolar dynamos as a function of Ra/Rac and Pm13 / 21
Introduction Modelling Axial dipoles Conclusion
Stability domain of the dipolar branch N% = 1.5
0 5 10 15 20 25 30 35
Ra/Rac
0
1
2
3
4
5
6
7
Pm
Dipolar dynamos as a function of Ra/Rac and Pm14 / 21
Introduction Modelling Axial dipoles Conclusion
Stability domain of the dipolar branch N% = 2.0
0 5 10 15 20 25 30 35
Ra/Rac
0
1
2
3
4
5
6
7
Pm
Dipolar dynamos as a function of Ra/Rac and Pm15 / 21
Introduction Modelling Axial dipoles Conclusion
Stability domain of the dipolar branch N% = 2.5
0 5 10 15 20 25 30 35
Ra/Rac
0
1
2
3
4
5
6
7
Pm
Dipolar dynamos as a function of Ra/Rac and Pm16 / 21
Introduction Modelling Axial dipoles Conclusion
Critical magnetic Reynolds number
∀N% , Rmc ∼ 102
0.0 0.5 1.0 1.5 2.0 2.5N%
101
102
103
Rm
0 5 10 15 20 25 30
Ra/Rac
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
N%
17 / 21
Introduction Modelling Axial dipoles Conclusion
Role of inertia: local Rossby number
N% = 0.5 N% = 1.5 N% = 2.0
0.02 0.04 0.06 0.08 0.10 0.12 0.14Ro`
0.0
0.2
0.4
0.6
0.8
1.0
f dipax
0.02 0.04 0.06 0.08 0.10 0.12 0.14Ro`
0.0
0.2
0.4
0.6
0.8
1.0
f dipax
0.02 0.04 0.06 0.08 0.10 0.12 0.14 0.16Ro`
0.0
0.2
0.4
0.6
0.8
1.0
f dipax
Ro` =Roc `c/π
as a function of radius for
dipolar (solid lines) and
multipolar (dashed lines)
dynamos at (N% = 2.5,Pm = 2)
(thin lines) and (N% = 3,Pm = 4)
(thick lines).
0.4 0.6 0.8 1.0 1.2 1.4 1.6
r
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.10
0.11
0.12
Ro? l
Ra = 5.4× 106;N% = 2.5
Ra = 6.4× 106;N% = 2.5
Ra = 7.4× 106;N% = 2.5
Ra = 9.0× 106;N% = 2.5
Ra = 8.0× 106;N% = 3.0
Ra = 9.0× 106;N% = 3.0
18 / 21
Introduction Modelling Axial dipoles Conclusion
Role of inertia: local Rossby number
N% = 0.5 N% = 1.5 N% = 2.0
0.02 0.04 0.06 0.08 0.10 0.12 0.14Ro`
0.0
0.2
0.4
0.6
0.8
1.0
f dipax
0.02 0.04 0.06 0.08 0.10 0.12 0.14Ro`
0.0
0.2
0.4
0.6
0.8
1.0
f dipax
0.02 0.04 0.06 0.08 0.10 0.12 0.14 0.16Ro`
0.0
0.2
0.4
0.6
0.8
1.0
f dipax
Ro` =Roc `c/π
as a function of radius for
dipolar (solid lines) and
multipolar (dashed lines)
dynamos at (N% = 2.5,Pm = 2)
(thin lines) and (N% = 3,Pm = 4)
(thick lines).
0.4 0.6 0.8 1.0 1.2 1.4 1.6
r
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.10
0.11
0.12
Ro? l
Ra = 5.4× 106;N% = 2.5
Ra = 6.4× 106;N% = 2.5
Ra = 7.4× 106;N% = 2.5
Ra = 9.0× 106;N% = 2.5
Ra = 8.0× 106;N% = 3.0
Ra = 9.0× 106;N% = 3.0
18 / 21
Introduction Modelling Axial dipoles Conclusion
Snapshots
N% = 1.5,Pm = 0.75,Ra/Rac = 5 N% = 2.5,Pm = 2,Ra/Rac = 3.4B
r(r=
r o)
Vr
19 / 21
Introduction Modelling Axial dipoles Conclusion
Table of contents
1 Introduction
2 ModellingGoverning equationsSet upMethods
3 Axial dipoles
4 Conclusion
20 / 21
Introduction Modelling Axial dipoles Conclusion
ResultsDipolar solutions can be observed at high N%, provided high enoughPm are considered.The critical Rm of the dipolar branch seems scarcely affected by theincrease of N%.The higher N%, the faster is reached the critical Ro` above whichinertia causes the collapse of the dipole. This explains why dipolardynamos become confined in the parameter space.The scarcity of dipolar solutions for substantial N% would thus rathercome from the restriction of the parameter space being currentlyexplored (because of computational limitations), rather than anintrinsic modification of the dynamo mechanisms.
References
Schrinner, Petitdemange, Raynaud & Dormy, 2014, A&A, 564, A78Raynaud, Petitdemange & Dormy, 2014, A&A, 567, A107Raynaud, Petitdemange & Dormy, 2015, MNRAS, 448:2055–2065
21 / 21
top related