Reasons for data discrepancy between Statistical approach ... · • J. Abdallah Jr, J. Colgan,...
Transcript of Reasons for data discrepancy between Statistical approach ... · • J. Abdallah Jr, J. Colgan,...
National Research Center
“Kurchatov Institute”
Moscow, Russia
2-nd Tungsten meeting, Vienna, Oct.6-8,2014
Statistical Theory for Radiative-
Collisional Processes with Heavy Ions
in Plasmas
V.S. Lisitsa, A.V. Demura, M.B. Kadontsev, V.A. Shurygin
Numerical codes for atomic processes and radiative losses
calculations • Kinetic codes
• D.E. Post, R.V. Jensen, C.B. Tarter, W.H. Grasberger, W.A. Lokke, At. Data and Nucl. Data Tables 20, 397 (1977).
• V.M. Leonov, V.E. Zhogolev, Plasma Phys. Control. Fusion 47, 903 (2005).
• Complex codes: atomic structure +elementary processes +kinetics
• H.P. Summers, ADAS users manual, JET –IR 06 (Abingdon: JET Joint undertaking) (1994)
• D. Post, J. Abdallah, R. E. H. Clark, N. Putvinskaya, Physics of Plasmas 2, 2328 (1995)
• K. Asmussen, K.B. Fournier, J.M. Laming, J.F.Seely, R. Dux, W. Engelhardt, J.C. Fuchs, Asdex upgrade team, Nucl. Fusion 38, 967 (1998).
• R. Neu, R. Dux, A. Kallenbach, T. Putterich, M. Balden, J.C. Fuchs, A. Herrmann, C.F. Maggi, M. O’Mullane, R. Pugno, I. Radivojevic, V. Rohde, A.C.C. Sips, W. Suttrop, A. Whiteford, Asdex upgrade team, Nucl. Fusion 45, 209 (2005).
• T. Putterich, R. Neu, R. Dux, A.D. Whiteford, M.G. O’Mullane, H.P. Summers, Asdex upgrade team, Nucl. Fusion 50, 025012 (2010).
• J. Abdallah Jr, J. Colgan, R.E.H. Clark, C.J. Fontes, H.L Zhang, J. Phys. B 44, 075701 (2011)
Reasons for data discrepancy between
different codes
• There are no universal model of atomic processes for total temperature range : 1eV – 40 keV
• 1 keV < T < 40 keV – Coulomb-Born
• 100 eV < T < 1 keV- renormalized Born- scattering matrix
• 10 eV < T < 100 eV- distorted waves
• 1 eV < T < 10 eV - close coupling – number of states?
• Strong discrepancy in ionization balance description as well radiative losses
Statistical approach vs
detail calculations
• Analog: plasma oscillation – molecular
dynamics;
• Manipulation not by specific probabilities
(lines) but by envelopes (averaged over
configurations by statistical method);
• Statistical Atom = collection of classical
oscillators with local plasma frequencies
and oscillator strengths.
Statistical models
• Looking for scaling
•Atom as a system of classical oscillators
• Oscillator strength f_ij is expressed in terms of atomic electron density
•Local Plasma frequency method (LPF) – Brandt-Lundquist (1966) as a method for
atom response calculation on external actions: atomic transition frequencies are
equal to local plasma frequencies
Photoabsorption cross section s(w) is a functional from electron density
( ) 24iff n r r drp= × × ×
( ) ( ) ( )( ) ( )2 2 2 23 22 2
4( )
abs p
p
r r
n re ed r n r r r
d rmc mc
drw
ww
p ps w d w w p
w
=
æ öç ÷ç ÷
= - = × ×ç ÷ç ÷ç ÷è ø
ò
Local plasma frequency model (LPF) in Thomas-
Fermi statistical model
2
2 42
0
( , )
( , )3/
16 ( , )1
( , )
B L
ph
a
x qx
x qeu a u
Z c x q
x x q
ww
w
w
w w
ccw p
sw c
c
-¢æ ö
= = × × ×ç ÷è ø -
¢16
u a uZ c
where хw is determined by the solution of equation
3/41/2
2
( )128, /
9TF
a
xu x r r
Z x
ww w
w
cww p
æ öæ ö= = × =ç ÷ç ÷è ø è ø
c(хw , q) is the solution of TF equation for the ion with a nuclear charge
Z and ion charge qZ
3/2
1/20 0
0
( , ) 1, 0( , )
( , )) ,( , ) 0, ( , )
x q xx q
x q qx q x qx
x
cc
cc c
® ®ì = í ¢= = -ïî
E.Fermi equivalent photon (EPh)
method in electron excitation of
complex ions • The electric field of impact electron on the
ion is equal to the action of equivalent photon flux (Fermi equivalent photon method)
• Electron excitation = equivalent photon absorption (photoabsorption cross section)
• Photoabsorption cross section→
• Quasicontiuum spectra
Fermi EPh method for calculations of atomic excitation and
ionization processes
( ) 3/2
3/2
2
2
0
[ ( / )] 2
2 2 2( / )
Coulombeffua E
RyaZ s
e a T
Z ZdI Z Z Rydu e g s u
TRyn d
a
w ww
w w
¥- -
æ ö× ×ç ÷è ø
ì üï ï æ öé ù< >ï ï æ ö× × × × ×ç ÷ê úí ý ç ÷ç ÷æ ö× è øê úï ï ë ûè øç ÷ï ïè øî þ
ò2 2
du2 2
ò2 22 2
-
Radiative losses in statistical model =
absorption of EPh
2
/ 243
0
0
3/2
3/2
2
( , )
( , )3 1 2/ ( ) (2 ) ( )
16 6 ( , )1
( , )
2
2 2
Z
ssI Z Ry
s
abs e a
s
s s
effu
RyZ s
T
x qx
x qRyQ n a Ry Z ds s
T x q
x x q
Z Z Rydu e g s u
T
ccp
wp c
c
×
¥- -
æ ö× ×ç ÷è ø
¢æ ö= × × × × × × × × ×ç ÷ç ÷
è ø -¢
æ öé ùæ ö× × × × × ×ç ÷ê úç ÷ç ÷è øê úë ûè ø
ò
ò
Total losses need for the sum over ionization balance
/ 2
3 2
0 02
0
3/2
3/2
2
1 2/ ( ) (2 ) ( ) [ ( ) / ]
6
2
2 2
ZI Z Ry
abs e a photo
effu
RyZ s
T
c RyQ n a Ry Z ds s a
e T
Z Z Rydu e g s u
T
w sp
×
¥- -
æ ö× ×ç ÷è ø
æ ö= × × × × × × × ×ç ÷ç ÷
è ø
æ öé ùæ ö× × × × ×ç ÷ê úç ÷ç ÷è øê úë ûè ø
ò
ò
1 21 23 23 23 23 23 2æ öæ öæ ö1 21 2c Rc Rc R3 23 23 21 21 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 2
In the frame of LPF model
Ionization balance = Gaussian shape for W
0.00 0.05 0.10 0.15 0.2010
-5
10-4
10-3
10-2
10-1
100
4030201510Te=5eV
rela
tive a
bundance, arb
.un.
x = k / Z
0.0 0.1 0.2 0.3 0.4 0.510
-5
10-4
10-3
10-2
10-1
100
10003001003010
rela
tive a
bundance, arb
.un.
x = k / Z
Rates are taken from
K. Asmussen et al. Nuclear Fusion 1998, vol. 38, No 7,
p.967.
Comparison with numerical codes
Comparison with kinetic codes
100
101
102
103
104
105
1E-32
1E-31
1E-30
LPF
CA-LARGE
ADPAKZIMPUR
From Polevoy
EM
Radiative-Collisional processes
in Statistical Theory
• All atomic processes can be expressed in
terms of transition energies and oscillator
strengths determined by atomic plasma
parameters related to electron density
distribution inside atoms; it is the case for
DR expressed in terms of radiative and
autoionization decay rates connected with
the same parameters.
Electron Impact Ionization=
Photo-Ionization by Equivalent
Photon Fermi • The statistical theory takes into account in a
specific manner collective effects related
• to excitation-autoionization channels by
excitation of plasmons with frequencies
lager than ionization potential
Tests for W, Xe, Kr, Fe, Ar and U IMPACT ELECTRON IONIZATION RATES AND CROSS SECTIONS
W ionization cross sections W ionization cross sections
Xe ionization cross sections Kr ionization cross sections
Fe ionization cross sections Ar ionization cross sections
U ionization cross sections DR-rates:Burdgess+General
expression
( )( )
3/2 2
2
2exp exp
2
R A
DR
n R A
W W nE z
T T W W n n T
pa
æ öDæ ö æ ö= - ç ÷ç ÷ ç ÷ +è ø è ø è øå
DR rate for W-29+ DR rate for W-20+
Spectral distribution of
radiative losses –
quasicontinuum spectra of W-
MD
Spectral distribution of
photoabsorption cross
section=envelope of spectral lines
( ) ( )
( ) ( ) ( )( )
( ) ( )
( ) ( )( )
( )
3
2
22
4
4;
in n
n
p
p
ph p
g f
g n r r d r
e n rr
m
n rr r
c n r
ww w
w
w d w w
w d w w
pw
p ws w w w
= -
= -
=
= =¢
å
ò
Slater orbitals – more realistic case
• If plasma frequency is lager than its
maximum value coming from Slater
distribution – sharp cut of the spectra at
large frequencies (small wave lengths)
( ) ( ) ( )2 2
0 0 exp 2nn r n R r n r rk g= = -
0
2 , ;
;
ii iI I ionization potential
n normalization factor
outer shell parameter
g
k
= - -
- -
- - -
Quasicontinuum spectrum of W-22-W-27
ions near 50 A spectral range (compare
with modern modeling of LHD-spectra)
Peaks
correspond to
small shifts of
plasma
frequencies
for different
ions
5 6 70
1
2
3
4spectrum from PLT, T
e=750 eV
theory
(Te=700 eV, k=4.5, R=200)
inte
nsi
ty [
a.u]
wavelength [nm]
Quasicontinuun spectra from
Alcator-C Mod and ASDEX Upgrade
4.5 5.0 5.5
1
2
3
4
5
theory
(Wtot
, k = 4.5)
ASDEX Upgrade
Alcator-C Mod
spec
tral
rad
iance
[a.
u.]
wavelength [nm]
Conclusion
• 1. Statistical models for atomic processes: ionization cross sections and, rates DR rates are in good correspondence with detail calculations and experiments
• 2. Quasicontiuum spectra of heavy ions seems to be collective plasma effects of atomic shell – envelope of spectral lines arrays
• -3. Resume: the precision of statistical model for atomic processes and radiative losses of plasma with heavy ions is of the same magnitude as in detail calculations
Tungsten spectra recorded at
the LHD and comparison with
calculations
C S Harte1, C Suzuki2, T Kato2,
H A Sakaue2, D Kato2, K Sato2,
N Tamura2, S Sudo2, R
D’Arcy1, E Sokell1, J White1
and G O’Sullivan1
1 University College Dublin,
Belfield, Dublin 4, Ireland
2 National Institute for Fusion
Science, 322-6 Oroshi-cho, Toki
509-5292, Japan
J. PhysB,2010, 205004 (14pp)