PFC/JA-83-34
Alcator DCT: Heat Removal and Impurity Control
in a Proposed Long Pulse Tokamak
B. Lipschultz, D. B. Montgomery, P. A. Politzer
S. M. Wolfe
Plasma Fusion Center
Massachusetts Institute of Technology
Cambridge, 14A
02139
October 1983
This work was supported by the U.S. Department of Energy Contract
No. DE-AC02-78ET51013. Reproduction, translation, publication, use
and disposal, in whole or in part by or for the United States government is permitted.
By acceptance of this article, the publisher and/or recipient acknowledges the U.S. Government's right to retain a non-exclusive,
royalty-free license in and to any copyright covering this paper.
Alcator DCT: Heat Removal and Impurity Control in a Proposed
Long Pulse Tokamak.
B. LIPSCHULTZ, D. B. MONTGOMERY, P. A.
POLITZER, and S. M. WOLFE,
Plasma Fusion Center, M.I.T.
The most crucial requirement
system is
its
performance
Alcator DCT device
for any heat removal and impurity control
under
at M.I.T.
steady
is designed
state
loading.
The
proposed
to provide a near term test
of long pulse and steady state issues integral to the tokamak program.
Anticipated plasma
parameters
include
1014
e
3
cm-
, Te ~
3-10
keV,
pulse time > 1 minute, plasma currents > 1 Mamp, 1 < ellipticity < 1.6.
Superconducting toroidal (Nb
3
Sn) and poloidal (NbTi)
field coils as well
as cw RF heating sources allow steady state operation.
All day operation
using RF current drive is contemplated.
The design parameters most relevant to this discussion are:
age wall loading (-
.2Mw/m
2
),
particle pumping rate (~
1021 part/sec)
and average heat load on the limiter or divertor plates (We have
been
studying
three
configurations
which
might
2-3Mw/m 2 ).
allow
obtain these parameters while keeping the impurity level low:
limiters,
2)
internal
poloidal
3) external
poloidal
divertors
(INTOR).
Pumped
simplest of
these
implement.
On the
other hand
to
divertors
(e.g.
ASDEX,
aver-
PDX,
limiters
us
1) pumped
DIVA)
and
are
the
the physics
data
base for pumped limiters is small and inadequate for extrapolation
a reactor.
There
exists a much more extensive physics
to
to
and engineering
data base concerning divertors.
The difference between alternatives 2)
and 3) lies in the divertor coil size, location and ease of maintenance.
The disadvantages
of
2),
for a
reactor,
are
that
links the TF and cannot be protected properly from neutron damage.
alternative 3)
avoids
these
disadvantages,
it
requires
expensive divertor
coil to allow the currents needed.
three alternatives
will
be
presented
and
further
coil
the divertor
a
While
large
and
Designs for all
comparisons
made.
The most
crucial
control system
loading.
is
requirement
its
for any
heat
removal
under
steady
state
performance
and
impurity
reactor
level
The proposed Alcator DCT device at M.I.T. will provide a near
term test of long pulse and steady state issues integral to the tokamak
program.
The objectives of the proposed machine are the development of
plasma edge, impurity, shape and profile control techniques appropriate
to steady
state high power operation;
drive and
heating
techniques;
and
the optimization
the demonstration
fully superconducting PF and TF magnet system.
description of the physical machine
meters.
Three impurity
and
of
of RF current
an integrated
This paper will give a
a review of
achievable para-
control designs that are contemplated will be
described and compared.
I.
Machine description
Alcator DCT is a high field, superconducting tokamak with dimensions
and parameters
shown in Table I.
Nb 3 Sn cable in conduit,
The TF magnets are constructed with
are non-circular in shape,
and have a peak field
of 10 T at the magnet and 7 T at the plasma major radius.
The supercon-
ducting PF
The
system
uses
multi-strand
NbTi
conductor.
moderate
aspect ratio of this device allows an air core transformer providing a
flux swing of 35 volt-seconds.
of pulse
lengths
Each toroidal
many
and
times
poloidal
This is the only planned tokamak capable
the
classical
field coil
has
magnetic
its
diffusion
own separate
time.
dewar
as
well as an overall outer vacuum vessel providing common thermal insulation.
The plasma has a minor radius
currents of 1.4 MA.
edge and
The toroidal
.04% on axis.
of 40 cm,
elongation up to 1.6,
and
field ripple is 2.0% at the plasma
The dee-shaped vacuum vessel has vertical and
horizontal dimensions of 155 and 110 cm.
This size provides sufficient
11
- 2 room for the internal
There are are 48 vertical ports and 12 horizontal ports (30 x
device.
70 cm)
components accompanying a steady state divertor
good diagnostic
access
personnel
allowing
access.
as well as
to the vessel
These features
are illustrated
providing
in Figure 1.
The engineering for Alcator DCT is described in more detail elsewherel.
RF systems consisting of 4 MW at the lower hybrid and 8 MW at the
ion cyclotron
these sytems
frequencies
would
be
are
presently
upgraded
to
on
site
for -use
cw
at
M.I.T.
on Alcator
Each
DCT,
of
with
reduction of the available ICRF power to 5 MW.
II.
Achievable plasma parameters
Transport during
forms for
auxiliary
heating
well
as
been modelled
from PDX
thermal diffusivity derived
injection experiments 2 , as
has
an
ohmic
"L-mode"
scaling
using two
neutral
beam
to account
for
"H-mode" confinement.
Transport calculations
assuming
of
9 MW
RF power
at the
source
indicate Alcator DCT can be expected to achieve electron and ion temperatures in the 5 - 10 keV range for average densities of 1 - 2 x 10
These parameters,
together
lengths of 1 to 5 minutes,
with
the available
flux
swing
14
cm-
3
.
imply pulse
even with the most pessimistic of transport
assumptions.
The resulting
equilibria
exhibit marked finite
beta distortions,
making possible the study of the evolution and control of high pressure,
shaped equilibria
However,
these
for
studies
times
exceeding
the
can be performed at
current
diffusion
time.
or close to the stability
boundary only if the more favorable confinement is achieved, or if some
additional heating power is supplied.
The available lower hybrid heating can be used to drive current
as well as heat the plasma.
Predictions
using the Fisch-Karney theory
- 3 indicate steady state currents of
13
densities >
5 x 10
current and
pressure
3
cm-
.
Studies of
profiles
pulse lengths greater
1 MA could be driven at electron
1
with
the evolution
RF
current
drive
than the plasma L/R time.
Alcator DCT parameters
this
time is
of
the
and control
will
of
require
For the anticipated
order
of
100
sec.
In
addition, the demonstration of day-long pulses, which will be required
for a tokamak
reactor,
places
severe
demands
on heat and
particle
removal and impurity control systems.
III.
Particle and impurity control techniques
A long pulse machine will need particle pumping to allow control
over the plasma density.
It will also require techniques for minimi-
zing the generation of impurities due to thermal loading and sputtering and for preventing their subsequent entrance into the main plasma.
The RF power deposited in Alcator DCT causes a high value of heat (.2
MW/m
2
) to flow into the plasma edge.
to values
for
which
large
impurity
This power flux is comparable
densities
are
intense auxiliary heating in present day machines.
observed
during
We conclude that
on a long time scale a simple limiter can not perform both the tasks
of impurity and particle control.
a long
pulse
test
of
the
two
Alcator DCT is planned to provide
most
promising
techniques
limiters and poloidal divertors.
For the latter
two separate
possibilities:
internal
PDX 3 , ASDEX 4 )
and
an
an
external
poloidal
-
pumped
option we consider
poloidal
divertor
divertor
(INTOR
5
).
(e.g.
By this
we are
referring
coils.
The designs and parameters for the pumped limiter and the two
to
a
divertor configurations
divertor
are
produced
presented
by
in
the
coils
outside
following
the
TF
sections.
Pumped limiter
Much experimental
and theoretical
research
is presently
con-
- 4 -
the
cerned with
of
subject
limiters.
pumped
experiments 6 - 9
have produced encouraging
whether pumped
limiters will remove
scale tokamak
without
questionable.
While their ease of construction,
feel it
we
engineering viewpoint,
risky
on a large
efficiently
of
overabundance
to
pumped limiters
causes
coils of divertors,
results on particle pumping,
particles
an
producing
scale
small
Although
is
impurities
compared to the extra
from an
attractive
be
high
a long-pulse,
to build
heat flux tokamak with only this option for impurity control.
Our basic design for a pumped limiter is shown in
Figure 2a.
It
is fully toroidal with neutralizer plates located under the limiter at
12 of the 24 bottom vertical ports.
the entrance
of
single leading
ports
edge
different shapes,
limiter is
movable
particles entering
where
is
and
they
can be pumped.
flat
to
along
a major
scrape-off
the
discharges
of
reasons,
the
same
10% of all
Approximately
radius.
layer will pass
strike the neutralizer plate 1 0 .
edge and
For
has a
The limiter
plasma
for
allow
and major radii.
sizes
the
Neutrals are therefore created at
behind the leading
We estimate that between 10
and 50% of the particles striking the plate will then be pumped.
that each
of the vertical ports has a conductance
and that
neutral
(~10Lp)
pressures
will
be
similar
to
Given
of ~ 2 x 104 1/sec
present
day
results
we will need ~ 10,000 1/sec of steady state pumping around the
torus.
Internal poloidal divertor
The poloidal
ly well
understood
of particles
into
divertor
produced
in comparison
the
efficient than with a
region
pumped
with internal
is relativeThe
the pumped limiter.
to
near
coils
the
limiter.
pumping
The
ducts
is
exhaust
much
more
impurity generation point
- 5 is removed
from the main plasma
the average
ties to
energy per
accumulate
ical poloidal
of neutrals
ion and
created
than a pumped
impurity-ion
in the divertor
divertor
has
limiter.
exhaust
ports.
We are studying this idea.
flux
The coil set
ASDEX:
a
equal but
ting and
main
wound
The
for
neutralized
this divertor
divertor
coil
opposite current.
covered by
coil.
were
inside
vacuum
pair
defining the
of
nulling
divertor
and
cooled tiles
absorb
parts
charge exchange flux.
helium for
the
coils
the
for
currently being
in
TFTR
This
is
300 W/cm 2 )
and
used
the INTOR design
we
are
to
which
the
bellows cover plates.
load
at
pump
similar to that
of
coils
an
carrying
be superconduc-
neutralizer
to
the
the
plates,
main divertor
baffling
2b).
The
are
also
and
some part
localized
could
of
baffles
are
fed
attaching
erosion
shown
tiles
the
to
system
walls
of
covered
caused
pursuing
in
are
the
the
steady
insufficient
brazing,
is
through
tiles
limits
currently
study 5 :
possible configuration
material'
part
if
less
the
with
by
the
The water coolant for these tiles and
superconductors
watts/cm 2 .
pumped may be
The
Figure
the
heat
the fraction
matching
attached
are
(see
of
mechanical technique
10's of
are
chamber
divertor chamber
to
tiles,
plates,
plates
chamber.
impuri-
the prototyp-
a triplet
two
causes
overcome
These three coils
the
coolant-backed
on
is
and
be
reduces
Because
subsequently
problem could
of the
.
neutralizer
which are
This
friction
12
chamber
toroidal
there
Flux multiplication1 1
edge.
for
attached
3.
The
would
heat
load
cooled
be
to
(200-
indicated
copper wool.
also
The
substrate
application
alternatives
copper pins and
Figure
cooled
state
our
supports.
by
One
'sandwich
used
for
- 6 External poloidal divertor
The poloidal divertor produced with external coils is not as well
understood as
the
internal
device with geometry
expanded boundary
coils.
that the
The
poloidal
similar
13
experiment
results
to
from the
expanded
boundary
divertor.
this class
in
which
General
only
of divertors
divertor
Atomic
exhibits
The
coils
operating
is the
link
DIII
the
TF
and JAERI teams indicate
particle
pumping
control similar to the PDX and ASDEX divertors.
and
impurity
The external poloidal
divertor throat is much more 'open' than that of the internal divertor.
This could lead to a greater leakage of neutral hydrogen and impurities
back to the main plasma.
the open divertor,
Field lines spread out to a greater extent in
allowing easier design of neutralizer plates.
the plasma shape or position is varied,
flux deposition
profile
will
vary.
the x-point,
Our
When
and thus the heat
calculations
indicate
that
these variations are less than or of the same order as the XE variations
discussed in Section IV.
We are considering two possible neutralizer plate/baffle configurations for the external divertor.
the neutralizer
plate
is mounted directly
with the baffles below.
2d,
the
neutralizer
approximately two
plate
thirds
on the
and baffle
of
the heat
are
heat loads,
although
2c)
vacuum vessel
combined.
and particle
is
some
2c,
wall
shown in Figure
This
flux
is because
contacts
the
Any neutral created will not have a direct
sight path back to the main plasma.
configuration (Fig.
shown in Figure
In the second configuration,
plate near the separatrix.
line of
For the first,
not
an optimized
The
fact that the first
shape results
of this load will be
in higher
spread out because of
perpendicular transport across the separatrix in the 'divertor chamber'.
While this shape is not optimized for heat load,
it
is easier to con-
- 7 -
struct and support than the second.
It is interesting to note that the
heat load results for the configuration shown in Figure 2d imply that
one should
This case
design
is
for
the
somewhat
design, which
is
shortest
different
quite
decay
from the
sensitive
to
length
deemed
reasonable.
internal poloidal
the
actual
divertor
value
of
the
decay length.
IV.
Heat deposition calculations
To study
the heat
case plasma loading,
deposition profiles,
10 Mwatts at the plasma edge.
length in the scrape-off layer,
results, is 1.2 cm.
using this
value
we have
chosen the worst
The energy decay
XE, which has been scaled from Alcator C
We have designed the limiter and divertor plates
and
have
calculated
the
resulting
heat
deposition
4 mm.
if XE is different by
The heat flux parallel to a field line in the scrape-off layer is
calculated on the plasma
midplane
using the
above
information.
flux surface geometry is given by the output (Psi(R,z))
The heat
brium code.
limiter surface
flow
assuming
is
then mapped
constant
flow
Psi(R,z) we can calculate the vector
at the plate,
giving
the heat
on
of an equili-
to the divertor
a
flux
The
plate
surface.
or
Knowing
components of the magnetic field
deposition.
The results
for
the four
different geometries discussed above are shown in Figure 4.
For all cases we find the heat loads and thus the thermally derived
impurity sources
tries (4b,
4d)
load below
2 MW/m
manageable.
to be
reactor prototypical.
have been
2
.
Two
optimized for XE = 1.2
The variation
of the four
cm,
geome-
to keep the heat
of heat loads with changing XE is
- 8 -
V.
Edge plasma parameters
A one-dimensional transport code, described in detail elsewherel 4 ,
used
has been
the plasma
model
to
convective
conductive and
or
either magnetically
processes
transport
(sinks).
impurity sources
field
The
This
edge.
line
and
as well as neutral
is
geometry
limited
mechanically
includes
calculation
to
adaptable
discharges.
For
this
calculation, as for the heat deposition model, we assume that 10 Mwatts
This
of power is deposited uniformly in the plasma scrape-off region.
power is deposited directly by the RF and/or is transported out of the
main plasma.
Typical results for the open divertor case are shown in
Figure 5.
With variation in the plasma density the flux multiplication
effect is
seen
for all
divertor
temperature in
plate.
the
Since
divertor
this
chamber,
model
we
predicts
can
[(l/ni)
reduces
x (dni/dx)]Fl/(Ldiv/3)
<
1
density
the
calculate
accumulation parameter from impurity momentum balance
parameter
effect
This
which also reduces the sputtering
carried per ion,
the average energy
at the
configurations.
divertor
12
over the
an
.
and
impurity
We find this
complete
range
in edge conditions.
The parameters
divertor chamber,
for
the
edge
of
the
main plasma,
as
opposed
were roughly similar for all configurations,
impurity density and power to the edge were kept constant.
to
as the
The edge
temperature and electron density ranges were 75 - 100 eV and .8 - 5 x
1013 cm- 3 respectively.
sputtering regime
With these values Alcator DCT
would enter a
predicted
Edge
typical
of
that
for
reactors.
and
divertor chamber parameters are summarized in Table lb.
VI.
Summary & conclusions
The three
configurations
we have
considered
for
the
long pulse,
- 9 high heat
flux
poloidal divertor
find the heat
and
sputtering
a
pumped
divertor.
external poloidal
and
flux
are
tokamak
DCT
Alcator
main plasma
similar
edge,
to
For all
three
we
if neutralizer plates
The pumped limiter may adequately pump particles
is at the
internal
reactor prototypical.
regime to be
localized at pump ports prove successful.
limiter,
However, because the surface
we
limiters,
simple
pumped limiters are risky with respect to impurity control
15
.
believe
Divertors
The heat is spread out due to
seem more promising for several reasons.
the length and spreading of field lines; the impurity generation point
is removed from the main plasma edge;
pumping is more easily allowed;
energy
reducing the
per ion,
flux multiplication
reduces
the
average
physical sputtering
source;
and
impurities accumulate in the divertor
chamber due to flow of hydrogen ions to the plate.
as part of
divertor,
include a poloidal
We will therefore
DCT design.
the Alcator
We
feel that with this impurity control option the possibility of achievIn addition,
ing long pulses will be greatly enhanced.
the pumped
limiter
comparison of
design
as part
two
particle
these
of
and
the basic
machine,
divertor
'open'
chamber.
techniques.
The internal coil
However,
the
average
field line length is greater for the external coil design (Figure
This is particularly important for the region X/XE < 1,
2/3 of the edge heat flow.
chamber effects.
a
providing
control
impurity
Each divertor option presents some advantages.
design permits a less
we will retain
6).
which contains
The additional length enhances two divertor
The heat load spreads due to perpendicular transport
and an impurity generated at the divertor plate must travel further to
return to the plasma edge.
the external
found the
poloidal
These effects may balance the openness of
divertor.
two configurations
From
an
engineering
roughly equivalent.
viewpoint,
At present,
we
we are
- 10 -
pursuing the external poloidal divertor
option because
reactor relevance and less understood physics.
of its
greater
Techniques for winding
a superconducting coil inside the chamber are also being studied should
the 'open' divertor fail.
Acknowledgements
The authors would like to acknowledge the work of the rest of the
Alcator DCT design group and Alcator C support staff.
thank the
Argonne
reactor
study
This work was supported by the U.
group
S.
for
In addition, we
their helpful discussions.
Department of Energy Contract No.
DE-AC02-78ET51013.
References
1
Alcator DCT Design Group,
Report PFC/RR-83-18.
2
3
M.I.T.
Plasma
Fusion
Center
Research
S. Kaye, Transport Analysis Workshop, PPPL, March 1983.
D. Meade,
et al., in Controlled Fusion and Plasma
9th European Conf., Oxford, 1979) Vol. 1 (1979)91.
Physics
(Proc.
4
M. Keilhacker, et al., 8th International Conference on Plasma
Physics and Controlled Nuclear Fusion Research, Brussels,
Vol. II, p. 351.
1980,
5
U. S. FED-INTOR Activity and U. S. Contribution to the International Tokamak Reactor Phase 2-A Workshop.
6
D. Overskei, Phys. Rev. Letters 46, 177 (1982).
7S. Talmadge, et al., J. Nuclear Fusion 22, 1369, (1982).
8
R. Jacobsen, J. Nuclear Fusion 22, 277 (1982).
9
P. Mioduszewski, J. Nuclear Materials 111 & 112, 253 (1982).
10J. Brooks, private communication.
11
M. Petravic, et al., Phys. Rev. Letters, 48, 326 (1982).
12
N. Ohyabu, et al., GA-A16434.
13M. A. Mahdavi, et al., GA-A16334.
- 11 -
14B.
Lipschultz,
to
be published
as an
M.I.T.
Plasma
Fusion
Center
Report.
15
E. Marmar, et al., this meeting.
Figure Captions
Figure 1
Cross-section of Alcator DCT.
Figure 2
Geometries for a) Pumped Limiter, b) Internal Poloidal
Divertor, c) External Poloidal Divertor using vacuum
vessel wall for mounting neutralizer plate and d)
External poloidal divertor shaped to reduce the heat
deposition < 2MW/m2.
Figure 3
Tile mounting design.
Figure 4
Heat deposition profiles for the 4 geometries shown
in Figure 2.
The curves shown are: E = 1.6 cm (*"*);
XE = 1.2 cm (-)
and XE = .8 cm (---).
In cases
b) and d) the plates are shaped for XE = 1.2 cm, to
force Pdep < 2 Mwatts/m 2 and only one plate out of
two is shown.
Figure 5
Typical profiles of ne, Te, PTOT, PCOND, PCONV along
a field line for divertor geometry.
Figure 6
Comparison of field line length vs. X/XE on the plasma
midplane.
The pumped limiter has field line length
equal to the asymptotic divertor value for all X.
For X/XE
fX =
0
E
P (4(1)) dl = (2 /3 )PIN
Table la.
Physical Characteristics of Alcator DCT
Plasma major radius, Ro
2.0 m
Plasma minor radius
0.40 x 0.56 m
Vacuum vessel bore
1.1 x 1.56 m
Toroidal magnetic field (R = RO)
7.0 T
Toroidal field ripple on axis
.04%
Maximum field in conductor
10.0 T
TF conductor
Nb3Sn
Poloidal field conductor
NbTi
Total flux swing (R - RO)
35 Wb
Plasma Current (q = 3)
1.0 MA
Heating
LH 4 MW(cw)
ICRF
5 MW(cw)
8 MW (30 sec)
1.5 - 5, w/ohmic
Maximum discharge duration
(minutes)
c, w/LH current drive
Table 1b.
Plasma Scrape-off Characteristics
PEDGE
10 Mwatts
71E
1.2 ± .4 cm
nE
.8
-
5 x 10
Te
75
-
150 eV
Pin/Awall
.19
Mwatts
=M
PAVE ON DIVERTOR PLATE OR LIMITER
2 - 3
1 3
Mwatts
m=
CX POWER FROM MAIN PLASMA
1 Mwatt
TOTAL CX & RAD LOSSES IN DIVERTOR
.5 - 6 Mwatts
XMFP (IONIZATION OF C)
.5 - 5 cm
cm-3
W~-'
6 6
---------------
man=
It~"
~
Fig.
Fig. 2a
.8
.7
- - - - -
.4.
-. 1Cl
cr
.. .. .
.
-7.
\
.--5
*
W#
C-
-.
2
-
- co
-
-. 6
- -- - - --. *4
-
MAJOR RADIUS (METERS)
Fig. 2b
.7
.4I
C/)
.
wj
.2
......
1i
.......
.....
....
..
..
..
..
..
...
..
.
.....-....
...
..
......
...
......
Un
r.D
co
0
cnO
-C*
It
,
ul)
--
MAJOR RADIUS (METERS)
Lo
I-
co
IL.
~
N N~c c
Fig. 2c
. . . . . . .. ...
.........
uj
uj
V.
......................
......................
** ..........
uj
......
0
..........
. . . . . . . . . . .
. . . .
. . .
. . .
. .
. . .. . . . . . . . . . .
CC
Lu
..
.............
Ln
to
r-
CO
as
0
C
t-4
pq
C'4
C,4
w
u-,
to
r-
'N
C4
C
MAJOR RADIUS (METERS)
Fig. 2d
.8
.7
.6
.4
Cl)
.3....
.
I*-
.. . ...
W
"N
- .. .... . .. . .. . .
..
. .......
LL
.
''
0
\IsJ\'.
.1
.....
.....
3.:
..
...
.. .. . .. .
o....
. ..
.. . .
-.
..
...
8j
*/......
tilI
--.
. .
.
. .
. .
..
%
MAJOR RADIUS (METERS)
J
.
BRK.E
...
Fig .3
....... ...
...
X :%
.....
.....
.......
.....
.....
.....
..........
.....
.....
........... ........
. ............
..........
...........
.....
..........
..........
........
.. .............
....
............
...
.. . . . . . . . .
...
.
.....
....... ...:. ... ...**
.....
..........
..........
.....
e
.................
.........................................
................
. . . . . . . . . . . . . ..
. . . ....
.. .. .. ..
..........
. .. ... .. . .. .. .
.....
.. ......
.
.. .
..
..
....
..
..
..
... . .. .
.
. . ... .. .. .
. .. ...
.
.
...........
... . . ...
..........
...... .......
..................
........ .......................
..............
......... ....
....... ....
. .
. ..............
..........
...........
............ . . .
............X
......... .
..
........
................
.....
.....
.....
.....
.....
.... . . ..
...........
COPPER Ft' R 0 R
STN . ST EEL
. F1 BER
Fig.
3.5
Cmlq
30
2.5
-
-.
-
2.0'
cc
w
e
Ue
I-
-
I:
*
1.5
1.0
CL
a:
,*.--I*
0
'*
0.5
0
10.0
20.0
30.0
DISTANCE ALONG
40.0
50.0
PLATE IN CM.
60.0
4a
Fig. 4b
3.5
'-,
I
i
Seperatrix
3.0 1
0
e.
*
*
2.5
*
*
h
6-
*
*
*
0
*
.
.
S
S
S
S
5
S
2.0c-
S
S
S
S
S
S
-------'
1.5-
S
w
S
S
S
S
S
1.0
S
S
S
S
S
S.
S
S
S
S
0
M~
0.5t
tO.
S
5.
1
I
'
55
~
0
5.0
10.0
15.0
20.0
25.0
**4
30.0
DISTANCE ALONG PLATE IN CM.
35.0
Pig.
10
I
I
81-
Seperatrix
I
I
I
61-
I
I
*
C',
-I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
410
iL
I
21-
0
.**
I*0.
I
..
*
I
p
.
/
/
0.2
0.4
0.6
4c
0.8
DISTANCE ALONG PLATE IN M.
1.0
Fig.
3.0
2.5t
-
2.0
6-j
L
--
1.5
1.0
0c
0.
0.5
0
..-
5.0
-
-Seperatrix
10.0
DISTANCE
15.0
20.0
25.0
ALONG PLATE IN CM.
30.0
4d
Fig.
5
112.0
120 I
100
Te
80
8.0
'E
U
0
60
x
n-
H
C
4.0
40
20
Divertor
Main Plasma
0
500
1000
1500
2000
2500
1.0
3000
-10
3500
Radiation
Convection
z
0
U
0.5
CL
Divertor
Main Plasma
0
2500
3000
DISTANCE ALONG FIELD LINE
(cm)
500
1000
1500
2000
3500
Fig.
I
I
I
I
60
0
External Poloidal Divertor
I
Internal Poloidal Divertor
40
20
0
0
Distance from
1.0
Separatrix
2.0
( X/E)
6
2