Extension of collisionless discharge models for 1/47 application to - - PowerPoint PPT Presentation

extension of collisionless discharge models for
SMART_READER_LITE
LIVE PREVIEW

Extension of collisionless discharge models for 1/47 application to - - PowerPoint PPT Presentation

Extension of collisionless discharge models . . . Extension of collisionless discharge models for 1/47 application to fusion-relevant and general Introduction plasmas Overview of existing models Analytic- numerical Leon Kos method (


slide-1
SLIDE 1

Extension of collisionless discharge models . . . 1/47 Introduction Overview of existing models Analytic- numerical method (ε = 0) Extension of the theoretical model (ε > 0) Results Conclusion

Extension of collisionless discharge models for application to fusion-relevant and general plasmas

Leon Kos

University of Ljubljana Faculty of Mechanical Engineering LECAD laboratory

October 21, 2009 / PhD Thesis Defense

slide-2
SLIDE 2

Extension of collisionless discharge models . . . 2/47 Introduction

Plasma Applications Motivation and thesis aims

Overview of existing models Analytic- numerical method (ε = 0) Extension of the theoretical model (ε > 0) Results Conclusion

What is plasma?

Fourth state of the matter (fire).

Solids - have very strong intermolecular bonds. Liquid - molecules are tied together by loose strings. Gas - atoms bounce around freely in space. Plasma - ionized gas, electrons and ions are separately free Temperature is the average amount of kinetic energy per atom.

slide-3
SLIDE 3

Extension of collisionless discharge models . . . 3/47 Introduction

Plasma Applications Motivation and thesis aims

Overview of existing models Analytic- numerical method (ε = 0) Extension of the theoretical model (ε > 0) Results Conclusion

Plasma properties

Quasi-neutral (ni = ne). Thin sheath is observed at the wall (λD ≪ L). Exhibit collective motion - collisionless. Very conductive - can be shaped and confined by electro-magnetic forces.

slide-4
SLIDE 4

Extension of collisionless discharge models . . . 4/47 Introduction

Plasma Applications Motivation and thesis aims

Overview of existing models Analytic- numerical method (ε = 0) Extension of the theoretical model (ε > 0) Results Conclusion

Laboratory plasmas

Aparatus used for producing a plane symmetric positive column in argon showing the position of the probes. [from Harrison-Thompson, 1959]

slide-5
SLIDE 5

Extension of collisionless discharge models . . . 5/47 Introduction

Plasma Applications Motivation and thesis aims

Overview of existing models Analytic- numerical method (ε = 0) Extension of the theoretical model (ε > 0) Results Conclusion

Fusion

Tokamak - Joint European Torus

slide-6
SLIDE 6

Extension of collisionless discharge models . . . 6/47 Introduction

Plasma Applications Motivation and thesis aims

Overview of existing models Analytic- numerical method (ε = 0) Extension of the theoretical model (ε > 0) Results Conclusion

Particle tracing developed in LECAD

Toroidal and poloidal magnets

Figure: Particle trajectories by Erˇ zen et al.

slide-7
SLIDE 7

Extension of collisionless discharge models . . . 7/47 Introduction

Plasma Applications Motivation and thesis aims

Overview of existing models Analytic- numerical method (ε = 0) Extension of the theoretical model (ε > 0) Results Conclusion

Plasma diagnostics

slide-8
SLIDE 8

Extension of collisionless discharge models . . . 8/47 Introduction

Plasma Applications Motivation and thesis aims

Overview of existing models Analytic- numerical method (ε = 0) Extension of the theoretical model (ε > 0) Results Conclusion

Geometry

One dimensional model

Φ(x) x = 0 x = L x = −L Φ(x) (x, v) Φ(x′) (x′, v′) Φw x

Figure: The geometry and coordinate system.

Plane–parallel geometry Symmetric - we observe only one half We normalize problem to L = 1.

slide-9
SLIDE 9

Extension of collisionless discharge models . . . 9/47 Introduction

Plasma Applications Motivation and thesis aims

Overview of existing models Analytic- numerical method (ε = 0) Extension of the theoretical model (ε > 0) Results Conclusion

Motivation

Provide precise treatment of the sheath region to fluid codes (SOLPS, EDGE2D). No analytic-numeric kinetic code available for Tn > 0. Particle In Cell (PIC) methods are not enough precise and can’t simulate ε = 0 case. Existing ε = 0 models are limited in temperature range. No solution to ε > 0 kinetic model available. Can velocity distribution function be obtained from potential curves?

slide-10
SLIDE 10

Extension of collisionless discharge models . . . 10/47 Introduction

Plasma Applications Motivation and thesis aims

Overview of existing models Analytic- numerical method (ε = 0) Extension of the theoretical model (ε > 0) Results Conclusion

Thesis statement

The problem of a special integro-differential equations should be solved numerically without approximations to achieve an extended solution range applicable to fusion-relevant and general plasmas for an arbitrary ion temperature and arbitrary finite ε.

slide-11
SLIDE 11

Extension of collisionless discharge models . . . 11/47 Introduction

Plasma Applications Motivation and thesis aims

Overview of existing models Analytic- numerical method (ε = 0) Extension of the theoretical model (ε > 0) Results Conclusion

Methodology

In this thesis the author presents investigations and results with the following assumptions

The Poisson equation is employed in the whole discharge region. A two-scale approximation is obtained just within the limit

  • f the infinitely small Debye length in comparison with the

system length. The ion-source temperature can take an arbitrary value. The electron-neutral impact is considered as a ionization mechanism.

slide-12
SLIDE 12

Extension of collisionless discharge models . . . 12/47 Introduction Overview of existing models

Tonks- Langmuir model Bissell-Johnson model Scheuer- Emmert model

Analytic- numerical method (ε = 0) Extension of the theoretical model (ε > 0) Results Conclusion

Overview of existing models

Two-scale approximation

Φ X

plasma solution exact solution sheath solution

Figure: Symbolic picture illustrating the two-scale approximation.

Plasma solution - Tonks–Langmuir model Sheath solution - Bohm model Exact solution - plasma + sheath (Our extended model)

slide-13
SLIDE 13

Extension of collisionless discharge models . . . 13/47 Introduction Overview of existing models

Tonks- Langmuir model Bissell-Johnson model Scheuer- Emmert model

Analytic- numerical method (ε = 0) Extension of the theoretical model (ε > 0) Results Conclusion

Plasma parameters

1 The macroscopic neutrality ne = ni 2 Strong electric field is localized to distance λD with

λD ≪ L , ε ≡ λD/L (≪ 1) , (1) where λD =

  • ǫ0kTe

n0e2 , (2) is the Debye radius.

3 The number of the particles in the Debye sphere is high

nλ3

D ≫ 1 .

(3)

slide-14
SLIDE 14

Extension of collisionless discharge models . . . 14/47 Introduction Overview of existing models

Tonks- Langmuir model Bissell-Johnson model Scheuer- Emmert model

Analytic- numerical method (ε = 0) Extension of the theoretical model (ε > 0) Results Conclusion

Tonks-Langmuir (T&L) model

0.0 0.2 0.4 0.6 0.8 1.0
  • 1.0
  • 0.8
  • 0.6
  • 0.4
  • 0.2
0.0 x T n =0 (T&L m
  • del)
s = -0.85403

Ions are born at rest (cold ion-source case). Analytic kinetic solution for ε = 0. Beakdown of quasi-neutrality at Φs = −0.85403.

Lewi Tonks and Irving Langmuir. A general theory of the plasma of an arc.

  • Phys. Rev., 34(6):876–922, Sep 1929.
slide-15
SLIDE 15

Extension of collisionless discharge models . . . 15/47 Introduction Overview of existing models

Tonks- Langmuir model Bissell-Johnson model Scheuer- Emmert model

Analytic- numerical method (ε = 0) Extension of the theoretical model (ε > 0) Results Conclusion

Bissell-Johnson (B&J) model (ε = 0)

  • 1.0
  • 0.5

0.0 0.5 1.0 2 4 6 8

  • 1.0
  • 0.5

0.0 0.5 1.0 2 4 6 8

θ θ F(θ) F(θ) τ = 0.5 τ = 0.33 τ = 0.25 τ = 2.0 τ = 1.0 τ = 8.0 τ = 4.0 τ = 0.1

Figure: Kernel F(θ) B&J equation (left, dashed), our approximation (right, dashed) and the exact kernel (solid).

Realistic Maxwellian ion-source velocity distribution. The Bohm criterion is used as the boundary condition to the quasi-neutrality equation. Kernel approximation with 8-th order Chebyshev polynomial and sinh(.) switch function. Plasma Eq. with 9-th order polynomial.

  • R. C. Bissell and P. C. Johnson.

The solution of the plasma equation in plane parallel geometry with a Maxwellian source. Physics of Fluids, 30(2):779–786, 3 1987.

slide-16
SLIDE 16

Extension of collisionless discharge models . . . 16/47 Introduction Overview of existing models

Tonks- Langmuir model Bissell-Johnson model Scheuer- Emmert model

Analytic- numerical method (ε = 0) Extension of the theoretical model (ε > 0) Results Conclusion

Scheuer-Emmert (S&E) model (ε = 0)

Better kernel approximation. Did not apply any kind of Bohm criterion in advance. Dense grid at endpoint singularity. Analytic approximation to sub-integrals with a series expansion. Different normalization than B&J. Ion source temperature range is still limited to non-fusion temperatures.

  • J. T. Scheuer and G. A. Emmert.

Sheath and presheath in a collisionless plasma with a maxwellian source. Physics of Fluids, 31(12):3645–3648, 1988.

slide-17
SLIDE 17

Extension of collisionless discharge models . . . 17/47 Introduction Overview of existing models

Tonks- Langmuir model Bissell-Johnson model Scheuer- Emmert model

Analytic- numerical method (ε = 0) Extension of the theoretical model (ε > 0) Results Conclusion

S&E and our results

What ion-source they employed?

Figure: Comparison of the potential profile with S&E for Ti = Te. The original scan is overlayed with our potential profile and axis box.

slide-18
SLIDE 18

Extension of collisionless discharge models . . . 18/47 Introduction Overview of existing models

Tonks- Langmuir model Bissell-Johnson model Scheuer- Emmert model

Analytic- numerical method (ε = 0) Extension of the theoretical model (ε > 0) Results Conclusion

OUR FUNDAMENTAL WORK

1 Analytic-numerical method (ε = 0) for wide temp. range 2 Extension of the theoretical model (ε > 0)

  • L. Kos, N. Jeli´

c, S. Kuhn, and J. Duhovnik. Extension of the Bissel-Johnson plasma-sheath model for application to fusion-relevant and general plasmas. Physics of Plasmas, 16(9):093503, 2009.

  • L. Kos, N. Jeli´

c, and J. Duhovnik. Modelling the plasma-sheath boundary for plasmas with warm-ion sources. In Proceedings of the International Conference Nuclear Energy for New Europe, pages 807.1–807.8, 2008.

  • L. Kos, J. Duhovnik, and N. Jeli´

c. Extension of collisionless discharge models for application to fusion-relevant and general plasmas, In NENE 2009, pages 820.1–820.10. 2009.

  • N. Jeli´

c, L. Kos, and D. D. Tskhakaya (sr.). The ionization length in plasmas with finite temperature ion sources. Physics of Plasmas, 2009. (under review).

  • M. Haefele, L. Kos, P. Navaro, and E. Sonnendr¨

ucker. Euforia integrated visualization. In PDP 2010 , Pisa, Italy, 2010. (accepted).

  • F. Castej´
  • n Maga na, L. Kos, et al.

EUFORIA: Grid and high performance computing at the service of fusion modelling.

  • Ibergrid. Grid Infrastructure Conference Proceedings, 12-14 May 2008, Porto, Portugal
slide-19
SLIDE 19

Extension of collisionless discharge models . . . 19/47 Introduction Overview of existing models Analytic- numerical method (ε = 0)

Discretization Stabilization of convergence Implementation aspects

Extension of the theoretical model (ε > 0) Results Conclusion

Analytic-Numerical method

Dimensionless quasi-neutrality equation

We are solving special integral equation in normalized form 1 B =

1

  • dx′ exp
  • ϑ +

1 2Tn

  • Φ(x′) −
  • 1 +

1 2Tn

  • Φ(x)
  • × K0

1 2Tn |Φ(x′) − Φ(x)|

  • (4)

Φ(x) is the unknown electrostatic potential B is the unknown constant which we fix by chooosing Φ(0) = 0. K0(z) is the modified Bessel function with logarithmic singularity at |z| = 0. Tn (neutral-gas temp.) and ϑ (ionization mechanism) are free parameters.

slide-20
SLIDE 20

Extension of collisionless discharge models . . . 20/47 Introduction Overview of existing models Analytic- numerical method (ε = 0)

Discretization Stabilization of convergence Implementation aspects

Extension of the theoretical model (ε > 0) Results Conclusion

Computational domain in 1-D

We introduce the following node positions for N points of the system

xi =

  • 1 − [1 − i/(N − 1)]λ2λ1 ,

i = 0, 1, . . . , N − 1 , (5) where λ1 and λ2 control the density at each boundary.

0.9999992 0.9999994 0.9999996 0.9999998 1.0000000
  • 0.80995
  • 0.80990
  • 0.80985
  • 0.80980
  • 0.80975
  • 0.80970
  • 0.80965
x

Figure: Last 28 points of the potential profile for Tn = 0.1, ϑ = 1 with N = 2401 points and grid density λ1 = 1, λ2 = 2.4.

slide-21
SLIDE 21

Extension of collisionless discharge models . . . 21/47 Introduction Overview of existing models Analytic- numerical method (ε = 0)

Discretization Stabilization of convergence Implementation aspects

Extension of the theoretical model (ε > 0) Results Conclusion

Discretized version suitable for iteration

exp

  • (1 +

1 2Tn )Vk

  • = B

N−1

  • i=0

xi+1

  • xi

dx′ exp[(ϑ + 1 2Tn )V (x′)] × K0 1 2Tn |V (x′) − Vk|

  • .

(6) Iterative formula (7) that evaluates to new Vk is mathematically exact, but can only be applied when all Vk are perfectly accurate. Vk = 1 1 +

1 2Tn

ln(B

N−1

  • i=0

Li), (7)

slide-22
SLIDE 22

Extension of collisionless discharge models . . . 22/47 Introduction Overview of existing models Analytic- numerical method (ε = 0)

Discretization Stabilization of convergence Implementation aspects

Extension of the theoretical model (ε > 0) Results Conclusion

What about eigenvalue B?

Only one equation (4) and two unknowns Φ(x) and B!

B appears to be a true eigenvalue of the system. B contributes to shift only. B can be calculated at any position like Eq. (7) during iterations.

200 400 600 800 1000 0.3004 0.3006 0.3008 0.3010

slide-23
SLIDE 23

Extension of collisionless discharge models . . . 23/47 Introduction Overview of existing models Analytic- numerical method (ε = 0)

Discretization Stabilization of convergence Implementation aspects

Extension of the theoretical model (ε > 0) Results Conclusion

Iterate using soft (time) step

V new

k

= Vk + α(Vl − Vk) (8) With sufficiently low α Eq. (6) converges! α averages many previous solutions. Practical values in range [0.0001, 0.1] Consequence - huge number of iteration steps required

slide-24
SLIDE 24

Extension of collisionless discharge models . . . 24/47 Introduction Overview of existing models Analytic- numerical method (ε = 0)

Discretization Stabilization of convergence Implementation aspects

Extension of the theoretical model (ε > 0) Results Conclusion

Can we speedup convergence somehow?

Yes, with parabolic interpolation near x = 0!

Vk = ax2

k + bxk + c ,

k = 0, 1, . . . , m (9) a = Vl − Vm x2

l − x2 m

, b = 0 , c = x2

l Vm − x2 mVl

x2

l − x2 m

, where mesh point xl is chosen at l = 3/4m. Practical value for the length of rewrite is from 1% to 10%. Completely disable it when approaching saturated solution.

slide-25
SLIDE 25

Extension of collisionless discharge models . . . 25/47 Introduction Overview of existing models Analytic- numerical method (ε = 0)

Discretization Stabilization of convergence Implementation aspects

Extension of the theoretical model (ε > 0) Results Conclusion

Solution smoothing

What is this good for?

A simple Laplacian-like smoothing technique with smooth-step parameter β V new

k

= Vk+β Vk−1 + Vk+1 2 − Vk

  • ,

k = N−1, N−2, . . . , 1 . (10) Prevents low frequency oscillations of the solution. Helpful for Tn ≤ 0.05. Practical range [0, 1]. Should vanish for the final solution.

slide-26
SLIDE 26

Extension of collisionless discharge models . . . 26/47 Introduction Overview of existing models Analytic- numerical method (ε = 0)

Discretization Stabilization of convergence Implementation aspects

Extension of the theoretical model (ε > 0) Results Conclusion

Implementation aspects

Direct integration using adaptive quadrature (QAG, QAGS) We extended Gnu Scientific Library (GSL) integration routines to 128 bit long double for improved accuracy. Parallelization using OpenMP standard Employment of XML schema for input Dump files for restarting Regrid for faster convergence from scratch

slide-27
SLIDE 27

Extension of collisionless discharge models . . . 27/47 Introduction Overview of existing models Analytic- numerical method (ε = 0)

Discretization Stabilization of convergence Implementation aspects

Extension of the theoretical model (ε > 0) Results Conclusion

Convergence demonstration for ε = 0

Tn = 10, iteration steps = 139000

  • 0.8
  • 0.7
  • 0.6
  • 0.5
  • 0.4
  • 0.3
  • 0.2
  • 0.1

0.2 0.4 0.6 0.8 1

Φ

x

Tn = 10, iteration = 0 ε = 0

slide-28
SLIDE 28

Extension of collisionless discharge models . . . 28/47 Introduction Overview of existing models Analytic- numerical method (ε = 0) Extension of the theoretical model (ε > 0) Results Conclusion

Extended model requires?

Simultaneous solving of

Boltzmann’s kinetic equation v ∂fi ∂x − e mi dΦ dx ∂fi ∂v = Si(x, v) , (11) with the ion-source term Si(x, v) Si(v, x) = Rnnne(x)fn v vTn

  • ,

(12) and Poisson’s equation − d2Φ dx2 = e ε0 (ni − ne) . (13)

slide-29
SLIDE 29

Extension of collisionless discharge models . . . 29/47 Introduction Overview of existing models Analytic- numerical method (ε = 0) Extension of the theoretical model (ε > 0) Results Conclusion

Extension of the theoretical model

Target equation for ε > 0

We are solving a special non-linear integro-differential equation with a singular kernel 1 B = 1 1 − exp(−Φ)ε2 d2Φ dx2 ×

1

  • dx′ exp
  • ϑ +

1 2Tn

  • Φ(x′) −
  • 1 +

1 2Tn

  • Φ(x)
  • × K0

1 2Tn |Φ(x′) − Φ(x)|

  • (14)
slide-30
SLIDE 30

Extension of collisionless discharge models . . . 30/47 Introduction Overview of existing models Analytic- numerical method (ε = 0) Extension of the theoretical model (ε > 0) Results Conclusion

Numerical method for ε > 0

Converted to relaxation method with Initial floating wall potential Φ[i] = Φw 1 − exp(1)

  • 1 − exp

i N

  • ,

(15) from ε = 0 case Floating wall condition exp(Φw) = 2π me mi

  • Tn

Te B 1 dx′ exp[Φ(x′)] , (16) is smoothly adjusted after converged state is reached.

slide-31
SLIDE 31

Extension of collisionless discharge models . . . 31/47 Introduction Overview of existing models Analytic- numerical method (ε = 0) Extension of the theoretical model (ε > 0) Results Conclusion

Convergence demonstration for finite ε

Tn = 0.1, ε = 0.001, iteration steps = 172000

  • 4
  • 3.5
  • 3
  • 2.5
  • 2
  • 1.5
  • 1
  • 0.5

0.2 0.4 0.6 0.8 1

Φ

x

Tn = 0.1, iteration = 0 ε = 0.001 ε = 0

slide-32
SLIDE 32

Extension of collisionless discharge models . . . 32/47 Introduction Overview of existing models Analytic- numerical method (ε = 0) Extension of the theoretical model (ε > 0) Results

The two-scale limit ε = 0 Unified plasma and sheath solution ε > 0

Conclusion

Potential profiles for ε = 0

0.0 0.2 0.4 0.6 0.8 1.0
  • 0.8
  • 0.6
  • 0.4
  • 0.2
0.0 T n 75.0 33.3 15.0 7.0 4.0 3.0 2.0 1.5 1.0 0.5 0.1 x

Figure: Potential profiles for various ion-source temperatures as

  • btained by us with the exact kernel (solid lines) and by Bissell and

Johnson with their approximate kernel (scattered).

slide-33
SLIDE 33

Extension of collisionless discharge models . . . 33/47 Introduction Overview of existing models Analytic- numerical method (ε = 0) Extension of the theoretical model (ε > 0) Results

The two-scale limit ε = 0 Unified plasma and sheath solution ε > 0

Conclusion

Plasma sheath boundary potential Φs

1 2 3 4 0.2 0.4 0.6 0.8 1.0 Scheuer & Em m ert's data scanned (from their Fig 2) Present work
  • b
=
  • s
T n /T e 20 40 60 80 100 0.2 0.4 0.6 0.8 1.0 Results of the present work
  • btained and presented in an
extrem ly wide range of ion source tem peratures
  • b
=
  • s
T n /T e (b) (a)

The plasma sheath boundary potential in a limited range of ion source temperatures, where the S&E approximate kernel is valid, in comparison with our results (a), and in a wide range of of the ion source temperatures (b), where we employed the exact kernel.

slide-34
SLIDE 34

Extension of collisionless discharge models . . . 34/47 Introduction Overview of existing models Analytic- numerical method (ε = 0) Extension of the theoretical model (ε > 0) Results

The two-scale limit ε = 0 Unified plasma and sheath solution ε > 0

Conclusion

Wall potential Φw

2 4 6 8 10 12 14 16 18 20
  • 3.6
  • 3.4
  • 3.2
  • 3.0
  • 2.8
  • 2.6
  • 2.4
T n w 0.00 0.25 0.50 0.75 1.00 1.25 1.50 B (b) (a)

The dependence of B (a) and of the wall potential (b) on the ion source temperature.

slide-35
SLIDE 35

Extension of collisionless discharge models . . . 35/47 Introduction Overview of existing models Analytic- numerical method (ε = 0) Extension of the theoretical model (ε > 0) Results

The two-scale limit ε = 0 Unified plasma and sheath solution ε > 0

Conclusion

Trajectory method

Method of characteristics for Velocity Distribution Function fi(Φ(x), v)

  • 1.0
  • 0.8
  • 0.6
  • 0.4
  • 0.2
0.0 0.2 0.4 0.6 0.8 1.0
  • 0.8
  • 0.6
  • 0.4
  • 0.2
0.0 0.2 0.4 0.6 0.8 t sx- t sx- t sx+ P P 2 P' 2 P T P 4 (x,v) P' 4 (x',v') t 4 t sx+ t 2+ t 3 t 2- t 1 v x x( )
  • 3.0
  • 2.5
  • 2.0
  • 1.5
  • 1.0
  • 0.5
0.0 x( ) (x) (b) (a)
  • 0.8
  • 0.6
  • 0.4
  • 0.2
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1 2 3 4 5 6 (c) f i f i f i v
  • 0.8
  • 0.6
  • 0.4
  • 0.2
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1 2 3 4 5 6 (f) 1 2 3 4 5 6 (e) x 1 2 3 4 5 6 (d) x 1 2 3 4 5 6 x (b) 1 2 3 4 5 6 x (a) v
slide-36
SLIDE 36

Extension of collisionless discharge models . . . 36/47 Introduction Overview of existing models Analytic- numerical method (ε = 0) Extension of the theoretical model (ε > 0) Results

The two-scale limit ε = 0 Unified plasma and sheath solution ε > 0

Conclusion

Velocity distributions

  • 1.0
  • 0.5
0.0 0.5 1.0 1.5 2.0 2 4 6 8 10 (b) T n = 10 T e f i v (x)
  • 0.01
  • 0.05
  • 0.1
  • 0.2
  • 0.3
  • 0.391
  • 1
  • 2
  • 3
2 4 6 8 10 (x)
  • 0.01
  • 0.05
  • 0.1
  • 0.2
  • 0.3
  • 0.813
  • 1
  • 2
  • 3
f i T n = 0.1 T e (a)

Velocity distributions for (a) Tn = 0.1, and (b) Tn = 10.

slide-37
SLIDE 37

Extension of collisionless discharge models . . . 37/47 Introduction Overview of existing models Analytic- numerical method (ε = 0) Extension of the theoretical model (ε > 0) Results

The two-scale limit ε = 0 Unified plasma and sheath solution ε > 0

Conclusion

Moments

  • 1.0
  • 0.8
  • 0.6
  • 0.4
  • 0.2
0.0 T n = 0.025 T n = 1.0 T n = 10 ln(n i , n e ) (a)
  • 1.0
  • 0.8
  • 0.6
  • 0.4
  • 0.2
0.0 0.0 0.2 0.4 0.6 0.8 1.0 (b) T n = 0.025 T n = 1.0 T n = 10 i = n i u i

(a) Ion and electron densities ni(Φ(x)) = ∞

−∞

fi(v)dv in a logarithmic presentation as a function of local potential Φ(x). (b)The ion flux Γi(Φ(x)) = ∞

−∞

vfi(v)dv as a function of Φ(x) .

slide-38
SLIDE 38

Extension of collisionless discharge models . . . 38/47 Introduction Overview of existing models Analytic- numerical method (ε = 0) Extension of the theoretical model (ε > 0) Results

The two-scale limit ε = 0 Unified plasma and sheath solution ε > 0

Conclusion

Effective ion (final) temperature Ti

  • 1.0
  • 0.8
  • 0.6
  • 0.4
  • 0.2
0.0 0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 0.16 0.18 0.20 (b) T n = 0.0 (Riemann limit T n =0) T n = 0.025 T n = 0.5 T n = 1.0 T i T n = 1.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 T i T n = 50 33.3 20 15 10 7 4 3 2 1 0.5 0.2 0.1 0.025 (a)

Profiles of the ion temperature Ti for various ion source

  • temperatures. The ion total

energy: Ki(Φ(x))) = 1 ni ∞

−∞

v2fi(v)dv Ion directional velocity: ui(Φ(x)) = 1 ni(Φ)Γi(Φ) The ion temperature: Ti(Φ(x)) = Ki(Φ) − u2

i (Φ)

slide-39
SLIDE 39

Extension of collisionless discharge models . . . 39/47 Introduction Overview of existing models Analytic- numerical method (ε = 0) Extension of the theoretical model (ε > 0) Results

The two-scale limit ε = 0 Unified plasma and sheath solution ε > 0

Conclusion

Ion temperature at the center and at the edge

10 20 30 40 50 1 2 3 4 5 6 7 T i,s (plasm a boundary) T i,0 (center of the system ) T i T n

Figure: The ion temperature at the center and the edge of plasma for various ion source neutral temperatures.

slide-40
SLIDE 40

Extension of collisionless discharge models . . . 40/47 Introduction Overview of existing models Analytic- numerical method (ε = 0) Extension of the theoretical model (ε > 0) Results

The two-scale limit ε = 0 Unified plasma and sheath solution ε > 0

Conclusion

Ionization length

2 4 6 8 10 12 14 16 18 20 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 H&T lim it L i =0.4046 L i = B(2 T n ) 1/2 S i =const S i ~ e T n H&T lim it L i =0.3444

Figure: Ionization lengths of the Maxwellian-source and flat-source ionization mechanisms as defined by H&T.

  • E. R. Harrison and W. B. Thompson.

The low pressure plane symmetric discharge. Proceedings of the Physical Society, 74(2):145–152, 1959.

slide-41
SLIDE 41

Extension of collisionless discharge models . . . 41/47 Introduction Overview of existing models Analytic- numerical method (ε = 0) Extension of the theoretical model (ε > 0) Results

The two-scale limit ε = 0 Unified plasma and sheath solution ε > 0

Conclusion

Potential profiles for various ε

0.0 0.2 0.4 0.6 0.8 1.0
  • 3.5
  • 3.0
  • 2.5
  • 2.0
  • 1.5
  • 1.0
  • 0.5
0.0 (c) T n = 10 = 0.0001 0.0003 0.0006 0.001 0.003 0.006 0.01 0.03 0.06 0.1 x
  • 3.5
  • 3.0
  • 2.5
  • 2.0
  • 1.5
  • 1.0
  • 0.5
0.0 (b) T n = 1.0 = 0.0001 0.0003 0.0006 0.001 0.003 0.006 0.01 0.03 0.06 0.1
  • 3.5
  • 3.0
  • 2.5
  • 2.0
  • 1.5
  • 1.0
  • 0.5
0.0 = 0.0001 0.0003 0.0006 0.001 0.003 0.006 0.01 0.03 0.06 0.1 T n = 0.1 (a)

Potential profiles for various ε and (a) Tn = 0.1, (b) Tn = 1.0, (c) Tn = 10.0. Our results obtained with the fixed system length L = 1. Wall potential Φw is dependent on Tn. Hydrogen was used for this simulation.

slide-42
SLIDE 42

Extension of collisionless discharge models . . . 42/47 Introduction Overview of existing models Analytic- numerical method (ε = 0) Extension of the theoretical model (ε > 0) Results

The two-scale limit ε = 0 Unified plasma and sheath solution ε > 0

Conclusion

Rescaled potential profiles for various ε

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6
  • 3.5
  • 3.0
  • 2.5
  • 2.0
  • 1.5
  • 1.0
  • 0.5
0.0 T n = 10 (c) = 0.0001 0.0003 0.0006 0.001 0.003 0.006 0.01 0.03 0.06 0.1 x
  • 3.5
  • 3.0
  • 2.5
  • 2.0
  • 1.5
  • 1.0
  • 0.5
0.0 T n = 1.0 (b) = 0.0001 0.0003 0.0006 0.001 0.003 0.006 0.01 0.03 0.06 0.1
  • 3.5
  • 3.0
  • 2.5
  • 2.0
  • 1.5
  • 1.0
  • 0.5
0.0 T n = 0.1 = 0.0001 0.0003 0.0006 0.001 0.003 0.006 0.01 0.03 0.06 0.1 (a)

Potential profiles for various ε and (a) Tn = 0.1, (b) Tn = 1.0, (c) Tn = 10.0. Rescaled results according to xs = √ 2π√TnB. From (b) it can be observed that wall potential Φw also changes with ε.

K.-U. Riemann. Plasma-sheath transition in the kinetic Tonks-Langmuir model. Physics of Plasmas, 13(6):063508, 2006.

slide-43
SLIDE 43

Extension of collisionless discharge models . . . 43/47 Introduction Overview of existing models Analytic- numerical method (ε = 0) Extension of the theoretical model (ε > 0) Results

The two-scale limit ε = 0 Unified plasma and sheath solution ε > 0

Conclusion

Zoom of the potential profiles

0.90 0.92 0.94 0.96 0.98 1.00
  • 3.5
  • 3.0
  • 2.5
  • 2.0
  • 1.5
  • 1.0
  • 0.5
0.0 = 0.0001 0.0003 0.0006 0.001 0.003 0.006 0.01 x

Figure: Potential profiles for various ε and Tn = 1 with a zoomed x range shows high precision results with ε ≤ 0.0006.

slide-44
SLIDE 44

Extension of collisionless discharge models . . . 44/47 Introduction Overview of existing models Analytic- numerical method (ε = 0) Extension of the theoretical model (ε > 0) Results

The two-scale limit ε = 0 Unified plasma and sheath solution ε > 0

Conclusion

Dependence on Tn

0.0 0.2 0.4 0.6 0.8 1.0
  • 3
  • 2
  • 1
x T n = 0.1 1 2 4 6 7 10

Figure: Potential profiles for various ion-source temperatures and fixed ε = 0.01.

slide-45
SLIDE 45

Extension of collisionless discharge models . . . 45/47 Introduction Overview of existing models Analytic- numerical method (ε = 0) Extension of the theoretical model (ε > 0) Results

The two-scale limit ε = 0 Unified plasma and sheath solution ε > 0

Conclusion

Charge imbalance

.0 .2 .4 .6 .8 1 .0 .0 .0 4 .0 8 .0 1 2 .0 1 6 .0 2 = 0.0001 0.0003 0.0006 0.001 0.003 0.006 0.01 = 0.1 0.03 0.06 (b) T n = 10 n i
  • n
e x .0 .0 4 .0 8 .0 1 2 .0 1 6 .0 2 = 0.0001 0.0003 0.0006 0.001 0.003 0.006 0.01 0.03 n i
  • n
e (a) T n = 1

Charge imbalance for (a) Tn = 1, (b) Tn = 10

slide-46
SLIDE 46

Extension of collisionless discharge models . . . 46/47 Introduction Overview of existing models Analytic- numerical method (ε = 0) Extension of the theoretical model (ε > 0) Results Conclusion

Our contributions

Extended temperature range with exact kernel. Derived quantities obtained from velocity distribution with direct integration. Extension to the case of finite ε. Future work

Precise investigation of the sheath edge singularity. Definition of PWT on the basis of VDF for ε > 0. Parallelization with MPI. Continuation of work in EUFORIA project.

slide-47
SLIDE 47

Extension of collisionless discharge models . . . 47/47 Introduction Overview of existing models Analytic- numerical method (ε = 0) Extension of the theoretical model (ε > 0) Results Conclusion

Singularity form for ε = 0 (preliminary results)

Finding power of α with fitting (Φs − Φ) → C(xs − x)α

0.01 0.1 1 10 0.50 0.51 0.52 0.53 0.54 0.55 0.56 0.57 0.58 0.59 0.60 0.61 0.62 0.63 0.64 0.65 0.66 0.67 0.68 max T n

Figure: Dependence of αmax on the ion-source temperature for logarithmic scale of Tn.