PFCJJA-88-17 L-Ray R. Plasma Fusion Center

advertisement
PFCJJA-88-17
L-Ray Diagnostic for the Compact Ignition Tokamak
R. D. Petrasso, C. L Fiore, and Chi-Kang Li
Plasma Fusion Center
Massachusetts Institute of Technology
Camhidge, MA 02139
April1988
Presented at
7 th
Topical Conference on High Temperature Plasma Diagnostics, Napa,
California, 13-17 March 1988. To be published in Review of Scientific Instruments.
y-Ray Diagnostic for the Compact Ignition Tokamak
R. D. Petrasso, C. L. Fiore, and Chi-Kang Li
Plasma Fusion Center, MIT, Cambridge, MA 02139
ABSTRACT
We have examined the feasibility of a fusion y-ray diagnostic for the Compact Ignition
Tokamak, which is expected to produce 1020 (d,t) fusion reactions per second. Gamma rays at 16.7
MeV are produced by the t(d,y)He 5 fusion reaction. A Monte-Carlo coupled neutron-photon
transport code is used to model the expected y-ray spectrum incident upon a Compton spectrometer
backed by a Cerenkov detector. The results indicate that a signal of -3 x 104 cts/s and a signal-tonoise ratio of -30 can be achieved.
INTRODUCTION
The measurement of the (d,t) fusion source function is essential for determination of burn
efficiency and a energy deposition within a hot plasma. The method most often discussed for
achieving this objective is based on measurement of the 14 MeV neutrons produced from the
t(d,n)He 4 reaction. Medley and coworkers( 1 ) pointed out, however, that energetic y rays from the
t(d,y)He 5 reaction provide an attractive alternative despite the small y branching ratio, 5 x 10-5.(2)
The feasibility of the y-ray approach is based on the premise that the 16.7 MeV y-ray signal is
measurable even in the presence of a large background of y rays resulting from (n,y) capture
reactions in the tokamak and surrounding structure. Moran( 3 ) and Ladish et. al.( 4 ) suggested
that a y detection scheme based on a Cerenkov threshold detector might be adequately sensitive to
2
the y signal while, at the same time, providing discrimination against the enormous y
background.
In this paper we semi-quantitatively evaluate such a diagnostic for the Compact
Ignition Tokamak, which will produce 1020 (d,t) fusion reactions per second.
EXPERIMENTAL DESCRIPTION
Figure 1 shows a simplified schematic of the plasma and y diagnostic, and Fig. 2 an
overview of the y detector. The plasma is viewed through a B1 0 /LiH plug, the function of which is to
partially attenuate source neutrons while transmitting source y rays. After passage through the
plug and the lead collimator, the y rays impinge upon a Compton convertor foil (2 mm of Be). One
16.7 MeV y ray in about 250 Compton converts within the foil, and of these approximately 1/3 are
forward directed within an -3' angle. (Compton electrons emitted at 3* (00) possess kinetic energy
of about 15 MeV (16.45 MeV).) These electrons next pass through a magnetic field of about 1500 G
and are deflected 90* into a Cerenkov C02 chamber. A huge background of low-energy Compton
electrons-from about .1 to 10 MeV-are also generated in the convertor foil. These electrons
correspond to the (n,y) background generated in the tokamak structure (especially in the inconel)
and in the plug, and they are largely prevented from reaching the Cerenkov chamber by the
magnetic screening field. Those electrons that enter the C02 chamber with kinetic energy in
excess of 14 MeV (corresponding to a C02 pressure of 1.4 atmosphere) produce Cerenkov
radiation.( 5 ) These photons are collected through a parabolic mirror( 3 ), and focused onto a wellshielded photomultiplier.
In order to make this design more quantitative, the signal and background y fluxes have
been evaluated using the MCNP (Monte Carlo coupled neutron photon transport code).( 6 ) The
materials and dimensions used as input to MCNP are approximately those depicted in Fig. 1.
Figure 3 shows the resultant y spectrum incident upon the converter foil. In addition, the total
neutron flux at the foil is about 300 times the y signal flux. Because of the low neutron sensitivity of
the Cerenkov chamber, it is felt that the neutrons will contribute negligibly to the noise level.
An estimate of the signal-to-noise ratio is obtained by using the overall sensitivity curve of
Brolley and Ladish et al.( 7 ) Multiplying together the y spectrum (Fig. 3), the Compton interaction
3
cross section( 8 ), and the Brolley/Ladish sensitivity curve, we obtain the curves shown in Fig. 4.
The integral under the solid (dotted) curve gives the signal (noise), resulting in S/N - 30. It is
important to note that without the screening magnet S/N
-
2. Thus the magnet is essential to
reduce the noise.
To estimate the signal, the following quantities are needed:
Fv,
fractional plasma volume viewed by the detector
-3 x 10-3
FQ,
fractional solid angle of the detector
-2 x 10-5
T,
y transmission through plug and inconel
-1/4
C,
Compton conversion factor (including forward scattering 5 3*)
-1/750
M,
electron transmission through the magnet into the Cerenkov chamber
-1/2
D,
detection efficiency in the Cerenkov chamber
~1/2
BR,
y branching ratio
5 x 10-5
Y-,
fusion neutron rate
10 2 0 /s
Signal (S):
S = (Fv)(Fa)(T)(C)(M)(D)(BR)(X)
-
3 x 104 counts/s.
CONCLUSIONS
In conclusion we find that a fusion y-ray diagnostic for CIT is an attractive possibility.
For a Cerenkov type detector with screening magnet, the signal-to-noise ratio and signal are of the
order ~30 and -3 x 104 counts/s, respectively.
ACKNOWLEDGEMENTS
This work is supported by DOE Contract # DE-AC02-78ET51013.
We have significantly
benefited from conversations with Drs. M. Moran, J. Ladish, S. Medley, and E. Cecil.
4
REFERENCES
(1) S. S. Medley et. al., Rev. Sci. Inst. 56 (5), 975 (1985); references therein.
(2) F. E. Cecil & F. J. Wilkinson, III, PRL 53 (8), 767 (1984); G. L. Morgan et. al. Phys. Rev.C. 33
(4), p. 1224 (1986).
(3) M. Moran, Rev. Sci. Intr. 56 (5), 1066, (1985).
(4) J. S. Ladish, et. al., IEEE Trans. on Nucl. Sc. 33 (1), p 385 (1986).
(5) V. P. Zrelov, Cerenkov Radiation in Isotropic and Anisotropic Media, 1968.
(6) MCNP: A General Monte Carlo Code for Neutron and Photon Transport, Version 3A, LA7396-M, Judith F. Briesmeister, Editor.
(7) J. Brolley, J. Ladish, et. al., SPIE 380, p. 256; Los Alamos National Laboratory, Conference on
Optics, 1983.
(8) J. H. Hubbell, Photon Cross Sections, from 10 keV to 100 GeV, NSRDS-NBS 29, August 1969.
5
wall
inconel
2 cm
inconel
Pb shielding
2
B10 LiH
\
Am
PMIfI~III~~,JI
' I I
I.Iin~m
ifff 11 If 1111
liii,,,
/ II
I
lilllilili
A'
I
I I If
8 cm
10 cm
20 cm
.d175 cm-
detector
I
-100 CM-4
50 cm
55 cm
340 cm
-130 cm- 1
wall
________________________a
77/
/ /
/ /
/
/
Fig. 1 Schematic of the CIT plasma and y-ray
diagnostic.
6
MMM
j
/I
/
/
neutron, y
dump
/
/
//
/
1 meter
A
2
Pb
//
chamber
e-
ax X
n
/
'I
19
x
x
x
x
X
X
K
1500 G
'U
PMT
2 mm
convertor
foil
//
Pb
y ray
Fig. 2
Schematic of the Detector. A small fraction
of the 16.7 Mev y rays Compton convert into
forward directed electrons. These high
energy electrons are bent via the magnetic
field into the Cerenkov chamber. Electrons
with energies greater than about 14 MeV
Cerenkov radiate for a C02 gas pressure of
1.4 Atm.
7
10-8
.
A.
..
1
1
1
1
II
I
.
I
I
I
I
I
I
I-~
I
I
L..
I
10- 9
~~1
g
I
I~~1
10-10
//
10-11
/
~~~1
-u
I
10-12
I--
I
L.
10-13
m
0
-
I
5
10
I
15
I I
,
IU
p
20
ENERGY (MeV)
Fig. 3 MCNP y spectra for the signal (solid line,
cross hatched) and the background (dotted
line). (Vertical units are relative.)
8
I
25
10-6
-
,
10-8
10-1210-14
10-12 -A---
0
5
10
15
20
ENERGY (MeV)
Fig. 4 The overall detection response curves for
signal (solid line, cross hatched) and
background (dashed line) y rays. The
integral of the solid (dashed) curve
corresponds to the signal (noise). (Vertical
units are relative.)
9
25
Download