IAEA Benchmark Experiment for Fusion Blanket Multiplier
NEUTRON MULTIPLICATION MEASUREMENTS IN BERYLLIUM, BERYLLIUM OXIDE
AND LEAD WITH 14MEV NEUTRONS
Tejen Kumar Basu
Bhabha Atomic Research Centre
Neutron Physics Division
Bombay  400 085, INDIA
Telefax: (9122) 5560750
Email: ntpd@magnum.barct1.ernet.in
I. Measured Quantities
The apparent leakage multiplication for 14MeV neutrons in 8, 12
and 20 cm thick beryllium metal, 20 cm thick beryllium oxide and 10
cm thick lead, all in rectangular geometry. This multiplication is
defined per source neutron and not per neutron entering the multiplier.
II. Method
The leakage neutron multiplication due to 14 MeV neutron in a
multiplier is defined as the total neutron production minus the
absorptions in the multiplier. This can be measured by using an
infinite 1/v moderating and absorbing medium surrounding the multi
plier. Similarly the source term can be measured without the multi
plier in place, and the ratio of the two measurements would give the
leakage multiplication. However, in practice a finite medium is used
and absorption measurements are made with a 1/v detector. In this case,
the ratio of absorption with and without the multiplier gives the
apparent multiplication. In these experiments, polyethylene/polypropy
lene was used as a 1/v absorber and BF3 detector was used to measure
the absorptions.
III. Geometry and Experimental Details
III.1. Beryllium metal assembly
This experiment was performed at the IRE, KFA, Juelich in
Germany along with V.R. Nargundkar, P. Cloth, D. Filges and
S. Taczanowski /1/. Beryllium metal rods of 1.85 g/cc density
having a 2 X 2 cm cross section and 65 cm length were used
for making the assembly. A 125 X 120 X 120 cm rectangular
polyethylene assembly was built with a 125 X 14 X 14 cm
central through channel to introduce the neutron generator tube.
The shape of the source channel was conical at the ends on either
side so that the source neutrons which do not contribute to multi
plication, leak out from the assembly. A central cavity was
provided to accomodate beryllium. For 12 cm thick beryllium, the
size of the central cavity was 65 X 38 X 38 cm and for 8 cm
and 20 cm thick beryllium, the size was 65 X 54 X 54 cm. Thus
in all the three assemblies, the outer dimensions of polyethylene
were kept unchanged.
12 cm beryllium assembly had 41 cm thick polyethylene in
the y and z directions and 30 cm thick polyethylene in the
xdirection. The 20 cm beryllium assembly had 33 cm thick poly
ethylene in the y and z directions and 30 cm thick polyethylene
in the xdirection. Since the beryllium stock was limited, the 20 cm
thick assembly was arranged in four configurations, details of which
are given in Ref./1/. The 8 cm beryllium assembly was also surroun
ded by 33 cm thick polyethylene in the y and z directions and
30 cm in the xdirection but the inner channel dimension was
65 X 38 X 38 cm whereas for 12 and 20 cm beryllium assembly,
the inner channel dimension was 65 X 14 X 14 cm. Details of the
dimensions are given in Ref./2/.
A 150kV Philips portable neutron generator tube capable of
giving a neutron yield of 10**9 n/s was used in the experiment.
The neutron generator tube was introduced at the centre of the
polyethylene assembly. The polyethylene assembly consisted of two
parts: the outer block surrounding the multiplier radially and the
end blocks surrounding it axially. The neutron absorption measure
ments were confined to a representative slab of the outer block by
symmetry considerations. The representative slab itself was further
divided into two parts: lower and upper. This slab had the flexibi
lity for creating a measuring channel along its length (xdirection)
at any height (z) and width (y). Except the representative slab where
measurements were made, rest of the polyethylene was fixed together by
means of polyethylene screws.
A 1cm diam, 5cm active length BF3 detector was used for the
measurement. A 2 X 2 cm axial channel was created in the block to
insert the detector. The measurements were made at seven values of z.
For each value of z, five channels were formed in the ydirection,
corresponding to 1, 6, 11, 21, and 31 cm. For each channel, seven
measurements were made in the xdirection. Thus about 250 points were
scanned in the representative block. The relative variation of the
neutron source intensity during individual runs was monitored by two
large BF3 detectors suitably located in the polyethylene assembly.
The monitor counts were normalised with regard to absolute neutron
yield, which was determined by the (n,2n) reaction in fluorine having
a threshold energy of 11.2 MeV. For this purpose, four 10mm diam,
2mm thick Teflon discs placed at 90 deg. apart on the target holder
were irradiated for 2 h.
The integration of the individual BF3 counts for the lower and
upper blocks of the representative slab was done using Simpson's
method. The exact geometry of the outer block of the polyethylene
assembly can be constructed by using 16 times the lower block plus
8 times the upper block. Therefore, the lower block integral counts
were given a weight of 16 and the upper block integral counts a
weight of 8. Then they were added together to obtain the net integral
counts in the outer block which is proportional to the net neutrons in
the polyethylene assembly. The measurement was done with and without
multiplier and the ratio of the neutron absorptions (integral counts)
in two cases give the apparent leakage multiplication.
III.2. Beryllium oxide assembly
In the case of beryllium metal, the neutron absorption measurement
in polyethylene was restricted to the outer block and not in the end
blocks. To see the effect of this, BeO (density=3.01 g/cc) measurement
was performed at Bhabha Atomic Research Centre,(BARC), Bombay using the
indegenously built 14MeV (d,t) neutron source. A 120 X 125 X 125 cm
near cubical polypropylene asembly was built with a 120 X 15 X 15 cm
axial through channel for introducing the neutron generator tube. The
20 cm thick BeO was surrounded by polypropylene of 35 cm radially and
30 cm axially /3/. The measurements in polypropylene were carried out
in representative blocks of both the outer block and the end block
regions. A 2 X 2 cm channel was created at a time and a 1cm diam
and 5cm active length BF3 detector was used to measure the neutron
absorptions at nine different axial locations. In all, 600 readings
were taken for the outer block and 100 readings for the end block.
III.3. Lead assembly
The leakage multiplication measurement in 10 cm thick lead assem
bly was also conducted at BARC., Bombay. The lead (density=11.24g/cc)
assembly had 45 cm thick polypropylene in the y and z direction
(radial) and 30 cm thick polypropylene in the x direction (axial)
surrounding it /4/. The neutron absorption measurements in polypropy
lene was done the similar way as in the case of BeO.
IV. Calculations
The Monte Carlo general geometry code MORSEE /5/ was used for
all calculations. The Los Alamos National Laboratory 30group CLAWIV
/6/ crosssection set in P3 scattering approximation and the response
functions, both based on ENDF/BIV data library were used in the cal
culations. The neutron source was placed at the centre of the assembly
and was assigned energy corresponding to the second group (13.515 MeV)
for isotropic source distribution. Calculations were also done for
anisotropic source but its effect was not much. B10 responses in poly
propylene were calculated with and without multiplier and the ratio of
these two values give the apparent calculated multiplication. Table I
summarises the experimental and calculated apparent leakage multipli
cation for Be, BeO and Pb. The values written in brackets are errors.
Table I
Experimental and calculated apparent leakage multiplication
_____________________________________________________________________

Geometry Description Apparent Multiplication
____________________________________________________________________
Multi Thick Outside Inside ChannelCalc. Expt. Expt/Cal
plier ness Dimension (cm)Dimension (cm)
_________________________________________________________________
   
Be metal 8 cm  65 X 54 X 54  65 X 38 X 38  1.64 1.33 0.81
   (0.01) (0.05)
Be metal12 cm  65 X 38 X 38  65 X 14 X 14  2.03 1.70 0.84
   (0.01) (0.05)
Be metal20 cm  65 X 54 X 54  65 X 14 X 14  2.07 1.68 0.81
   (0.01) (0.20)
   
BeO 20 cm  60 X 55 X 55  60 X 15 X 15  1.54 1.19 0.77
   (0.01) (0.05)
   
Pb 10 cm  60 X 35 X 35  60 X 15 X 15  1.55 1.53 0.99
   (0.01) (0.03)
_________________________________________________________________
V. Conclusion
For beryllium metal and beryllium oxide, the measured leakage
multiplication values are nearly 20% smaller than the corresponding
calculated values whereas for lead, the experimental leakage multi
plication agrees fairly well with the calculated value.
References
1. T.K. Basu, V.R. Nargundkar, P. Cloth, D. Filges and S. Taczanowski,
"Neutron Multiplication Studies in Beryllium for Fusion Reactor
Blankets", Nucl. Sci. Eng., 70, 309 (1979)
2. V.R. Nargundkar, T.K. Basu and O.P. Joneja, "Reanalysis of Neutron
Multiplication Measurements in Thick Beryllium and Graphite
Assemblies for 14MeV Neutrons", Fusion Technol.,12, 380 (1987)
3. V.R. Nargundkar, T.K. Basu, O.P. Joneja, M.R. Phiske and S.K.
Sadavarte, "Neutron Multiplication Measurement in BeO for 14MeV
Neutrons", Fusion Technol., 6, 93 (1984)
4. O.P. Joneja, V.R. Nargundkar and T.K. Basu, "14MeV Neutron Multi
plication Measurement in Lead", Fusion Technol., 12, 114 (1987)
5. "MORSEE, A General Purpose Monte Carlo Multigroup Neutron and
Gamma Ray Transport Code", CCC258, Radiation Shielding Informa
tion Center, Oak Ridge National Laboratory (1975)
6. R.J. Barrett and R.E. MacFarlane, "CLAWIV Coupled Neutron and
Photon Cross Sections for Transport Calculations", LA7808MS,
Los Alamos National Laboratory (1979)