PFC/RR-83-9
DOE/ET-51013-78
UC20
A PRELIMINARY BUNDLE DIVERTOR
DESIGN FOR ALCATOR DCT
T. F. Yang
A. Wan
P. Gierszewski
A preliminary design of an optional bundle divertor for Alcator DCT
has been obtained.
This report will briefly describe the advantages of
using bundle divertors and summarize the results from the DITE operation.
The preliminary magnetic and thermal hydraulic design of the DCT bundle
divertor will be presented along with suggestions for minor machine modifications to accommodate the divertor.
Finally, the reactor relevant aspect
of the bundle divertor option will be discussed.
1
1. ADVANTAGES OF BUNDLE DIVERTORS
There are the following advantages:
0
Plasma shape, vacuum chamber and TF coil can be circular and
the plasma can still be up-down symmetric
*
Target and pumping systems can be outside
0
Divertor separatrix is permanent.
Divertor operation is
independent of plasma control (beta, shape, etc.)
*
The effect on system design is minor
2
2. CONCLUSION FROM DITE EXPERIMENT
The experimental evidence from DITE can be summarized as follows:
*
Separatrix of the divertor acts as magnetic limiter
0
No experimental evidence of adverse effect on plasma confinement
0
No experimental evidence of adverse effect on ion confinement
during beam heating
*
Effects on the plasma
(a) Increases the central Te from 200 eV to 380 eV
(b) Reduces hard X-ray flux
(c)
N0 emission reduced by an order of magnitude
(d) Fe, Cr emission reduced 70%
(e) Particle exhaust is nearly 100%
(f) Ion temperature doubled by beam injection to T. = 610 eV
(g) 02 is reduced only 20% (contamination may occur during
startup)
3
..
.......
3. PRELIMINARY DESIGN AND OPERATIONAL EVALUATIONS
The toroidal flux configuration is shown in Fig. 1. Table 1 gives
the design parameters used for DCT and the preliminary bundle divertor
design parameters.
As shown in Fig. 1, there is plenty of space inside
the separatrix and between the TF coil and scrape-off edge for conductor
and structure.
Figure 2a shows a 3-D perspective view of the first (front)
coil of the cascade divertor configuration shown in Fig. 2b.
Four types of operational methods have been evaluated and listed in
Table 2. Steady state water cooling is possible, but requires 30 MW of
electric power.
Liquid nitrogen cooling will reduce the power to 3 MW, but
may need large quantities of liquid nitrogen.
reduce the power requirement to 1 MWe.
Gaseous helium cooling will
The average current density is
6.7 kAmp/cm 2 which gives 10 kAmp per turn of current for the same superconductor cable used for DCT.
The maximum field is estimated to be 8 T.
The stability of the conductor has to be studied.
for optimization.
However, there is room.
Current density can be reduced and the use of a super-
conductor is quite feasible.
Figure 3 compares the DITE MKlA and Alcator C divertors.
There are
the following improvements:
9
Null trace is bending toward the plasma and the plasma is less
distorted
*
The null trace is taller so that qD is larger and the efficiency
may be better
a
The plasma exhaust exit is larger
*
The ripple on axis is smaller
Similar results can be expected from the DCT divertor design.
4
4. SUGGESTIONS FOR MINOR MACHINE MODIFICATIONS FOR IMPROVEMENT
It is more feasible for the use of superconductors if the following
minor modifications can be made:
9
Make the TF coil cross-section rectangular with longer side
in the R direction
*
Reduce the number of coils to 22 if allowable in the toroidal
ripple.
*
These modifications will ,also increase the access spaces for
all purposes
*
The scrape-off layer thickness is now designed to be 7 cm.
The current density can be reduced immediately if 5 cm or
less of the scrape-off is used
5
5. REACTOR RELEVANT
The divertor configuration proposed for DCT has been proven reactor
relevant for INTOR-type.
0
The bundle divertor was designed as an option
0
It is designed as a plug-in device
0
There is 60 cm of shielding space
*
The overall current density is only 3.5 kAmp/cm 2 ; therefore it
can be superconducting
*
The divertor chamber is outside the tokamak system
*
A Monte Carlo modeling has shown that the magnetic surfaces
are well formed with island structure.
The thermal conductivity
is only 40% larger than the neoclassical value of axisymmetric
plasma.
The impurities diffuse outward.
6
TABLE 1
BUNDLE DIVERTOR DESIGN PARAMETERS
Ro = 2.0 m
BO = 7.0 T
a = 0.4 m
Ascrape-off
ID/coil = 800 kAmp
Cross-section = 0.10 m x 0.12 m
2
JD = 6.7 kAmp/cm
DIVERTOR OPERATION
Water cooled
30 MWe
Steady state
N2
Superconductor
3 MWe
Steady state
Bmax =8 T
I/turn = 10 kA
stability?
He
1 MW
e
Steady state
7
TABLE 2
DCT BUNDLE DIVERTOR THERMAL-HYDRAULICS
Coil:
10 cm x 12 cm
10 kA/turn
1.5 cm2 /turn
80 turns
0.9 cm2 conductor
0.34 cm2 coolant
Mass
flow
(hgls)
Pei
(MW)
Pref
(MW)
Liters
cool ant/
pulse
Tin(k)
Tmax
300
383
144
29
Liquid N2
70
97
144
4.2
49
10300
Gaseous He
30
56
1.2
47
6700
Normal Conductor
Water
*
14
single-phase, 10 atm inlet pressure, single-turn cooled.
7 Tesla, 100 RRR copper
Superconductor
Using coil similar to TF superconductor:
10 kA/turn in a peak field. of 7 T (?)
Compare with TF coil parameters
5800 kA/turn in 9 T operating
7950 kA/turn in 9 T critical current
8
FLUX LINES
.GO -
.40-
.20'j
(M)
IXIXIXI
.00
FX71XIX
-. 20-. 40-
-. Go-
.80I
2.0
I
I
2.2
2.4
2.6
2.8
3.0
3.2
X (Mv.)
Figure 1.
Toroidal flux configurations
9
It
is
at
n
LI
va
.0
wl
n
ri
n
0
It
0
I'
'4
5-
1"
~
EM
2U
W
W
U
~
~
-
-i
-
-
-
1
1
~
~U3
40
u2
.-
K
St
'4
L.
I-
r
WI
rl
rt
LO
CM
'a
vs
n
-CD
Z
Lt
Vt
61
U
N
tU
S
U
F
3F
10
-
1 !i
0)
0.
DITE
Modaus
B
SIOgnotoin cami
k-n lotflr orbil
Rs
16
41.2
R (m)1-1.
1.0
1.2
14
1.-.
ALCATOR
-
%
%%
~
t".
.
-
*
.
.
.
*
Figure 3. Comparison between DCT divertor
and MKA divertor.
11
I In
X
--
LIi
-o
4J
-J
0a
c
0)(
-rC0.
0
) 4J-
-Q
(WO)
-~
@.-.o
-8
a) X a)
z
v -S.-
-
,->
-D
-r=
-5-O
(0.0
0
(n
ILi
0
CL)
-ar
c
S
-.
-
90
0
*e.
tI
~I
. vi
oio
I*
(WO)
Z
12
5-
-
r._
0 0~-
0,09-
(Urc
v.0
-D
::,
A-I
o
0
S.-
4