Stable organized motion and coherent structures in fusion ... 2006... · Publications September...
Transcript of Stable organized motion and coherent structures in fusion ... 2006... · Publications September...
Stable organized motion and coherent structures in fusion
plasmas
• Introduction• Calculation of the spectrum in the problem
of perturbed zonal flow (similar to Orr-Sommerfeld) ; numerical description
• Continuous model for the stationary states of the Hasegawa-Mima equation
Plasma Theory Group
Members: F. SpineanuM. VladL. Anton (long term visit in UK)
Main subjects of interest :
Studies on instabilities, turbulence and transport in confined plasmasStatistical physics of test particle motionFirst principle models of plasma self-organised flows, coherent structures
Major recent results• Determination of the effects of intrinsic trajectory trapping or eddying
produced by the ExB drift in non-linear regimes.
Analytical methods for determining the statistical properties of test particles evolving in turbulent fields:
- DCT (Decorrelation Trajectory Method);- NSM (Nested Subensemble Method)
– Applications
• Field theoretical methodsA fundamental theoretical structure, largely extending the classical concepts in fluid/plasma description– Applications
PublicationsSince the beginning of Association (2000):
35 articles in international refereed journals40 International Conferences (Proceedings)
• Physical Review Letters : 6 articles• Physical Review : 8 articles• Physics of Plasmas : 2 articles• PPCF : 5 articles• Nuclear Fusion : 1 article
Average impact factor/person/year = 5.23
NOVA Science Publishers
(M.V. and F. S)
Publications September 2005 – September 20061. M. Vlad, F. Spineanu, S. Benkadda, “Impurity pinch from a ratchet process”, Physical Review Letters 96
(2006) 085001.2. F. Spineanu, M. Vlad, ´Statistical properties of an ensemble of vortices interacting with a turbulent field´,
Physics of Plasmas 12 (2005) 112303.3. M. Vlad, F. Spineanu, S. Benkadda, “Mass charge effects for impurity ratchet pinch”, in preparation (2006). 4. F. Spineanu, M. Vlad, “Density pinch from self-organizing vorticity in cilindrical plasma”, in preparation (2006).5. M. Vlad, F. Spineanu, “Turbulent pinch in non-homogeneous confining magnetic field”, ”, in preparation (2006).6. F. Spineanu, M. Vlad, K. Itoh, S.-I. Itoh, ‘Review of analytical treatments of barrier-type problems in plasma
theory’, Progress in chemical Research, Editor A. N. Linke, Nova Publishers (2006), ISBN: 1-59454-483-2.7. F. Spineanu, M. Vlad, ‘Soliton self-modulation of the turbulence amplitude and plasma rotation’, Progress in
Soliton Research, Editor L. V. Chen, Nova Publisher (2006), ISBN 1-59454-769-6.8. M. Vlad, F. Spineanu, “From Brownian diffusion to the non-linear transport by continuous movements”, in
‘Albert Einstein Century’, Editor J. Alimi, American Institute of Physics (2006) (in print).9. F. Spineanu, M. Vlad, “A background trend to ordered states in confined plasmas”, 11th European Fusion
Theory Conference, Aix en Provence, 26 September 2005.10. M. Vlad, F. Spineanu, “Test particles, test modes and self-consistent turbulence”, 11th European Fusion
Theory Conference, Aix en Provence, 26 September 2005.11. M. Vlad, F. Spineanu, P. Mantica and EFDA JET contributors, „Ratchet pinch and density peaking in H-mode
JET plasmas”, 11th EU-US Transport Task Force Workshop, Marseille 2006.12. F. Spineanu, M. Vlad, „The common ground of the density pinch and the transition to high confinement
regimes in tokamak”, 11th EU-US Transport Task Force Workshop, Marseille 2006.13. F. Spineanu, M. Vlad, „Field theoretical methods in theory of turbulence relaxation”, BG-URSI School and
Workshop on Waves and Turbulence Phenomena in Space Plasmas, 2006, Kiten, Bulgaria, invited paper.14. M. Vlad, F. Spineanu, „Nonlinear effects in charged particle transport in turbulent magnetic fields”, BG-URSI
School and Workshop on Waves and Turbulence Phenomena in Space Plasmas, 2006, Kiten, Bulgaria, invited paper.
15. F. Spineanu, M. Vlad, „Helicity fluctuation, generation of link number and effect on resistivity”, 21st IAEAFusion Energy Conference, Chengdu, China, 2006, accepted.
16. M. Vlad, F. Spineanu, „Test particle statistics and turbulence in magnetically confined plasmas”, 21st IAEAFusion Energy Conference, Chengdu, China, 2006, accepted.
Collaborations
CEA – Cadarache + Universite de Provence (Marseille)
ENEA - Frascati
ContactsK. Itoh and S.-I. Itoh (Japan)D. Montgomery (Dartmouth, USA)P. H. Diamond (San Diego, USA)K. Spatschek (Juelich, Germany)F. Jenko (Garching, Germany)
JET, Transport Task Force
Plasma rotation, density pinch, impurity inflow
Stable organised motion and coherent structures in fusion plasmas
Generation and stability of coherent flows in tokamak plasmasSimulation of alpha particle transport by collective modes
Stable organized motion and coherent structures in fusion plasmas
Most of the work in the period January – September 2006 was concentrated on the following milestones of our Work Plan:
• Assessment of the effect of the magnetic field inhomogeneity on test particle transport and calculation of the impurity pinch and diffusion coefficients in turbulent plasmas using the decorrelation trajectory method.
• Calculation of the spectrum in the problem of perturbed zonal flows (similar to the Orr-Sommerfeld); numerical description.
• Continuous model for the stationary states of the Hasegawa-Mima equation (small scale decay of the ITG turbulence).
We have also continued several studies concerning the transport coefficients in turbulent plasmas and we have initiated a new subject
1
Milestone : Calculation of the spectrum
in the problem of perturbed zonal flow
(similar to Orr-Sommerfeld); numerical description
Previous work : derivation of the exact analytic solution of the Flierl-
Petviashvili equation.
Present objectives:
• explain the streamers and intermittency of the ITB
• understand the role of the tilting instability
Target:
• JET experiments on ITB
• gyrokinetic simulations of sheared flow
• ITER-relevant : threshold for the ITB formation
– Cluj 2006 PTG Milestones I
2
Pattern of the zonal flow on
Jupiter (data from Voyager 1 and
2, cf. Busse, Chaos, 1994).
Suggested form (Hasegawa,
Kodama) and numerical
simulation
(compare L-H transition flows)
– Cluj 2006 PTG Milestones I
3
The first major problem: generation of ZF
Class A
1. Rayleigh-Benard analogy : Reynolds stress drives cell-flow transfor-
mation (Howard-Krishnamurti, Finn et al. for RB, Pogutse et al.
for interchange);
2. interaction of a set of essential spectral components (four-wave
paradigm) (Chen, White, Zonca);
3. thermodynamic view: wave-kinetic Fisher equation for k-space
pseudoparticles: predator-pray model (Diamond et al.);
Class B
Correlated growth of secondary instabilities of elongated ITG cells
Reynolds stress + convective cell transversal instability
(see numerical simulations of Kishimoto, Idomura et al.)
– Cluj 2006 PTG Milestones I
4
Possible scenario: reconnection of strongly tilted eddiesof the ITG mode
GyroFluid simulations
(Kishimoto)
The main driving mechanism
is Reynolds stress:
1. The tilting instability;
2. Gradual transformation:
ITG to electron drift;
3. The periodic layered
flow is an attractor;
(still to explain the super-fast
transition (∼ exp(exp(t)))
– Cluj 2006 PTG Milestones I
5
ITG versus Drift wave dynamics
∂∂t
(1 − ∇2
⊥
)φ −
[(−∇⊥φ × ez
)· ∇⊥
]∇2
⊥φ
+∇‖v‖i + vD
(1 +
1+ηiτ
∇2⊥
)∂φ∂y
= 0
∂v‖i∂t
+[(
−∇⊥φ × ez)· ∇⊥
]v‖i
= −∇‖p − ∇‖φ + µ∇2‖
v‖i
∂p∂t
+[(
−∇⊥φ × ez)· ∇⊥
]p
+vD1+ηi
τ∂p∂y
= −Υ∇‖v‖i
∂∂t
∇2⊥φ +
[(−∇⊥φ × ez
)· ∇⊥
]∇2
⊥φ
+κnτ
∂∂y
∇2⊥φ − µ∇2
⊥φ + ∇‖V‖e = 0
V‖e =mi
meη∇‖ (φ − p)
∂p∂t
+(−∇⊥φ × ez
)· ∇⊥p
−κn∂φ∂y
+ ∇‖V‖e = 0
– Cluj 2006 PTG Milestones I
6
The Flierl-Petviashvili equation
∆φ = αφ − βφ2 (1)
The exact solution
φ (x, y) =α
2β+ s℘
(iay + ibx + ω|g2 =
3α2
(sβ)2
)(2)
with the condition
a2 + b
2 =sβ
6(3)
– Cluj 2006 PTG Milestones I
7
Figure 1: Periodic solution of the Flierl-Petviashvili eq.
– Cluj 2006 PTG Milestones I
8
Comparison with experimental observations
(Doublet III-D, Coda et al., PRL2001)
Experiment Analytic solution φs
ϕrms > 10 V ϕrms > 17 V
ωE×B ∼ 2 × 105 s−1 ωE×B ∼ 2.2 × 105 s−1
λr ∈ (15...30) ρs λr ≃ 17.4ρs
– Cluj 2006 PTG Milestones I
9
−10 −8 −6 −4 −2 0 2 4 6 8 100
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
0.5
kx (cm−1)
S(k
x)
The spectrum of the stationary potential perturbation
Figure 2: Experimental and theoretical spectrum.
– Cluj 2006 PTG Milestones I
10
Comparison with numerical simulations
(Gyrokinetic, Lausanne group, Alfey et al., Varenna, 2002)
Simulation Analytic solution φs
Er ∼ (−24 · · · + 24) × 103 (V/m) Er ∼ (−23 · · · + 23) × 103 (V/m)
ωE×B ∼ 2 × 10−3 Ω−1 ωE×B ∼ 10 × 10−3 Ω−1
δx ∼ (8.1 · · · 13.5) ρs δx ≃ 13.6ρs
– Cluj 2006 PTG Milestones I
11
The second major problem: destruction of ZF
Stability and structural stability
Figure 3: A weak and large scale perturbation evolves to a system of
vortices.
– Cluj 2006 PTG Milestones I
12
The stability of sheared flows
Typical instabilities
1. Kelvin-Helmholtz instability; cat’s eye Kelvin-Stuart (Reynolds con-
dition: existance of an inflection point in the velocity profile)
2. dissipative laminar sheared flow: Orr-Sommerfeld (our first order
from the barotropic eq.)
3. Kolmogorov flow: perturbation on a spatially periodic flow : chaos
– Cluj 2006 PTG Milestones I
13
The linear stability of the equation with scalar
nonlinearity
The equation of drift waves with the presence of a temperature gradient
∂
∂t
(1 −∇2
⊥
)φ =
∂
∂y
[(−∇2
⊥ + 4η2)φ]− φ
∂φ
∂y
The stationary solution is perturbed by a small function ε (x, y, t)
φ → φs (x, y) + ε (x, y, t)
where φs (x, y) is the periodic flow solution (expressed in terms of the
Weierstrass function). We obtain
∂
∂t(1 − ∆) ε =
∂
∂y
(−∆ε + 4η
2ε − φsε −
1
2ε2
)
We look for the linear dispersion relation.
– Cluj 2006 PTG Milestones I
14
The dispersion relation
∂2ε
∂x2+
iky
iω + iky
(− ω
ky− ωky − k
2y − 4η
2 + φs
)ε = 0
Notations q1 ≡ 1
1+ ωky
and a1 ≡ − ωky
− ωky − k2y − 4η2
∂2ε
∂x2+ (q1a1 + q1φs) ε = 0
φs (x, y) = φ0 + s℘ (x)
where φ0 = 3α2β
. The equation becomes
∂2ε
∂x2+ [q1a1 + q1φ0 + q1s℘ (x)] ε = 0 (4)
– Cluj 2006 PTG Milestones I
15
The Mathieu equation is an approximation
1
ky=
ω ± Q1/2
P + 1
1 ∓
[1 +
2 (P + 1)
(−ω + Q1/2)2
]1/2 (5)
1
ky≃ ω ∓√
Q
P + 1
(1 ∓
√2√Q
)
We obtain λ⊥ = 3ρs and from numerical simulations, λ⊥ = 6ρs
A necessary alternative: transversal stability of the flow in the
elongated ITG cells
– Cluj 2006 PTG Milestones I
16
Conclusions to this milestone The stationary sheared flow discovered
by us previsously as a solution of the Flierl-Petviashvili equation is very
close to experiment and simulation. The formation of such a structure is
connected with the tilting instability and reconnection of eddies aligned
along the poloidal direction.
The stability shows first destruction and generation of cells of the order
few ρs.
– Cluj 2006 PTG Milestones I
17
Future workAnalyticFind the description of the late phase of the tilting instability and
of the reconnection of eddies (Eckhaus instability).
NumericTime dependent simulations of of tilting and reconnection of eddies.
– Cluj 2006 PTG Milestones I
18
The collaborationsJETIn Transport Task Force. Pedestal physics (Andreev, etc.)
FrascatiLimit cycle of low order system for the quasi-periodic destruction
of the flow (Zonca, Annibaldi, Briguglio, Gregorio Vlad)
– Cluj 2006 PTG Milestones I
1
Milestone : Continuous model for the stationary states
of the Hasegawa-Mima equation
Previous work : derivation of the sinh-Poisson equation for the Euler
fluid.
Present objectives:
• explain the density pinch in tokamak and generation of background
gradients
• explain the low- to high- confinement transition
Target:
• JET experiments on density pinch
• DIII-D experiments on poloidal velocity profile at LH transition
• ITPA scaling for ITER
– PTG Milestones II
2
First part : What is behind the
evolution of vorticity to coherent
structures
Euler equation. D. Montgomery et
al., Phys. Fluids A4 (1992) 3.
– PTG Milestones II
3
There is a fundamental dynamics of the ideal plasma/neutral fluid leading
to evolution toward regular structures and coherent flows. It exists in the
background of the instabilities and turbulence.
It is justified to say that the vorticity self-organises
Two systems related with plasma/fluid vorticity elements exhibit
self-organization
• The model of point-like vortices interacting via a potential allows the
representation of the vorticity as a self-organising fluid, which acts as
a drive for passive fields like density. It has a negative temperature
(Taylor, Edwards);
• the continuum limit of the system of point-like vortices can be
described as the equations of continuity and momentum conservation
for an ideal compressible fluid; this fluid has negative pressure
– PTG Milestones II
4
There are two fundamental models :
• the line evolving under Nambu-Goto action leads to a fluid with
negative pressure and strange polytropic (Chaplygin gas).
• a fluid/plasma with vorticity is equivalent with a system of point-
like vortices interacting in plane via a potential
– long-range (Coulombian, logarithmic) : Euler fluid. No density
(ρ = const) The stationary states sinh-Poisson equation. Exactly
integrable.
– short-range (screened, K0) : Charney-Hasegawa-Mima
plasma/atmosphere. The third direction is implicitely present.
The density is not constant.
How are they connected: both reduce to the sinh-Poisson equation, along
two ways: at self-duality and as condition of emdedding in R3 according
to Gauss-Codazzi eqs.
– PTG Milestones II
5
The fact that the density follows the vorticity is due to the existence of an
intrinsic finite length in 2D plasma: the Larmor radius
The Charney-Hasegawa-Mima equationThe equation (CHM) derived for the two-dimensional plasma drift
waves and for Rossby waves in meteorology is:
(1 −∇
2
⊥
) ∂φ∂t
− κ∂φ
∂y− [(−∇⊥φ× n) · ∇⊥]∇2
⊥φ = 0 (1)
This is the equation governing the stationary states of the CHM eq.
∆ψ +1
2p2sinhψ (coshψ − p) = 0
– PTG Milestones II
6
Second part. How the vorticity evolves to a co-
herent structure and how the density responds
The conservation laws are less useful here. We need dynamical equations
• the aggregation-coagulation statistical process (Phase I, at the
beginning of the discharge when a random distribution of vortices is
generated. The merging takes place at a rate t−3/4 and the result is a
central vortex and possibly a ring of opposite vorticity).
• the equations of motion of the field theoretical model (Phase II)
• the equations of the ideal fluid with negative pressure
• the Manton’s method : dynamics on the parameter manifold, close to
self-duality
– PTG Milestones II
7
L = −κεµνρtr
(∂µAνAρ +
2
3AµAνAρ
)(2)
−tr[(Dµ
φ)† (Dµφ)]
−V(φ, φ
†
)
Sixth order potential
V(φ, φ
†
)=
1
4κ2tr
[([[φ, φ
†
], φ
]− v
2φ)†([[
φ, φ†
], φ
]− v
2φ)]
. (3)
The Euler Lagrange equations are
DµDµφ =
∂V
∂φ†(4)
−κενµρ
Fµρ = iJν (5)
– PTG Milestones II
8
The density reaction is via the effective Larmor radius
• The vorticity builds up a central cluster surrounded by a ring of
vorticity of opposite sign. The density must follow the vorticity
(Ertel’s theorem). This is only possible due to the polarisation drift,
compressible, which implies (but does not explicitely uses) the third
dimension.
• The pinch of vorticity is at the origin of the pinch of density; the
density develops gradients and in consequence, diamagnetic flow
• The combination of diamagnetic flow vd and of rotation speed u
induces a change in a basic parameter: ρs is replaced by an effective
Larmor radius1
(ρ
effs
)2=
1
ρ2s
(1 − vd
u
)
• When the density dragged by the vorticity develops higher gradients,
– PTG Milestones II
9
the effective Larmor radius becomes very large ρeff → ∞. The model
of point-like vortices with short-range interaction (Hasegawa-Mima)
evolves to the point-like vortices with long-range interaction (Euler
fluid). The density decouples from the vorticity;
– PTG Milestones II
10
Confirmation that our equation is the correct description :
Numerical solution for L = 307 : mono- and multipolar vortex
– PTG Milestones II
11
Third part. The space of solutions andhow the pinch arises.
The numerical study of the equation reveals the existance of three type
of states
• smooth vortex, stable
• strongly localised (narrow) vortex, a physical approximation of the
singular vortex
• a class of intermediate quasi-solutions, organized along a line of
minimum departure from the action extremum.
The smooth vortices are accessible from a wide range of initial conditions.
The narrow vortices are only accessible from a subset of initial functions.
There are differences: for narrow vortices the maximal vorticity is much
higher compared to the smooth vortex. There is a difference in both
energy and vorticity between them.
– PTG Milestones II
12
– PTG Milestones II
13
The H (high) confinement state as coherent
structureCoherent structure (Lagrangian) vs Driven-dissipative (spectrum).
0 20 40 60 80 100 120 140 160−0.1
−0.05
0
0.05
0.1
0.15
0.2
0.25
radius
ψ s
trea
mfu
nctio
n
The solution streamfunction ψ(r)
Figure 1: Natural edge sheared flow in TJ-II Stellarator (Alonso et al.)
and our solution for the potential.
– PTG Milestones II
14
Comparison with the self-organisation of the vorticity obtained in a
statistical approach (HW)
Figure 2: A theory of self-organization of drift turbulence (Hasegawa-
Wakatani) leads to this radial profile for the potential. Actually their
model does not explicitely need excitation of drift waves.
– PTG Milestones II
15
Edge electric field in DIII and JET
– PTG Milestones II
16
−400 −300 −200 −100 0 100 200 300 400−8
−6
−4
−2
0
2
4
6
8
10
12x 10
−3
r
vψ
The magnitude of the velocity vector tangent to the streamlines
Results for the poloidal velocity
– PTG Milestones II
17
– PTG Milestones II
18
Transition from smooth to narrow requires an input of kinetic en-
ergy and localised vorticity
−20
−10
0
10
20
−10−5
05
1015
0
100
200
300
400
ln(Efin
)
ln(|Ωfin
|)
L
Figure 3: The squares are smooth and the dots are narrow vortices. The
plot presents the difference in both kinetic energy and vorticity. At right is
a qs. for L = 401 with slow intermediate variation of vθ.
– PTG Milestones II
19
Scatterplot (Energy, Vorticity) of final results for pinch and for
sheared velocity vθ
0 0.001 0.002 0.003 0.004 0.005 0.006 0.007 0.008 0.009 0.01−1
−0.5
0
0.5
Final energy
Fin
al v
ortic
ity Ω
Small dots: all data; green squares: evolution; red circle: narrow; black square: solution
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1−20
−10
0
10
20
30
40
50
60
70
Final energy
Fin
al v
ortic
ity Ω
Small dots: all data; green squares: evolution; red circle: narrow; black square: solution
Figure 4: Black squares: smooth solutions; green: intermediate quasi-
solutions; red: singular solutions. In addition, for barriers: yellow: stable
qs. with ring; blue: intermediate forms evolving to stable ring qs.
– PTG Milestones II
20
The role of the effective Larmor radius
vd/u ρeff
s/ρs L = a/ρeff
s|vθbottom|
∣∣∣vphys
θbottom
∣∣∣ (m/s) |Er| (kV/m)
0.2 1.118 273 1 × 10−3 979 2.45
0.4 1.29 236 1 × 10−3 979 2.45
0.85 2.58 95 2.2 × 10−3 2152 5.38
0.95 4.47 63 0.02 19580 48.95
– PTG Milestones II
21
Conclusions to this milestone As part of the phenomenology of struc-
ture generation in fluids and plasma it appears convenient to take the
vorticity as the active factor of self-organization. This is supported by
the model of point-like vortices interacting in plane, model that has pro-
vided us with the description of the coherent stationary states. Examin-
ing the solutions of the coherent vortical type and their neighborhood in
the function space it is found that there are states with properties similar
to the physical states of 2D plasma. The pinch of density is the evolution
of the system along a string of quasi-solutions, between a smooth vortex
and a singular vortex. The increase of the effective Larmor radius due
to the increase of the density gradient slows down this process without
suppressing it. The transition between the L to H states appears as the
transition between the smooth vortex and a state with high vθ shear (a
quasi-solution which is close to the extremum of action), of which it is
separated by a gap in energy and vorticity.
– PTG Milestones II
22
Future workAnalyticProve that intermediate profiles are softly unstable. Find the equa-
tion of evolution of vorticity close to the absolute extremum (singu-
lar vortex), via Manton’s method. Derive a unique chain of equa-
tions for density pinch, from Ertel’s theorem and effective ρs.
NumericTime dependent simulations of CHM.
– PTG Milestones II
23
The collaborationsJETIn Transport Task Force (Mantica, Naulin, Tala, Rassmussen,
Parail, etc.). Contacts with experiments (Andrew) for poloidal
rotation, quite recent.
CEA-Marseille and ENS-ParisMuch better time dependent simulations using Adaptive Mesh Re-
finement and comparison with our stationary solutions (Kai, Farge,
Benkadda, Agullo, etc.)
2006 : Two invited researchers in the last year (from Frascati and
from CEA-Marseille) One month mobility for a member of our group,
in France.
– PTG Milestones II
24
Interesting theoretical development (spin-off)
We are able to provide an accurate description of the
large-scale atmospheric vortex : tropical cyclone (ty-
phoon, hurricane)This is confirmed by quantitative comparisons with ob-served profiles of azimuthal velocity, radius of the eye-wall, decay of velocity.
– PTG Milestones II