§ 
5 
é 
5 
= 
2 
oo 
oO 
8 
Counting Rate (cps) 
FIG. 3. Source moves with detector, at fixed distance below detec- 
tor. Distances below are as follows: for curve A-2.0 cm; for curve 
B-6.2 cm; for curve C-10.8 cm 
beryllium. Safe distance from this 
source is 56 em for continuous expo- 
sure, based on a maximum dose of 20 
fast neutrons/cm?/sec., 
The counter could be moved verti- 
cally just outside the tank wall, carry- 
ing a centimeter scale. As the scale 
traveled past a fixed pointer, counter 
position was indicated to +0.2 mm. 
Two types of neutron counters used. 
One type of neutron counter used was 
a boron trifluoride counter 2.5 cm in 
diameter containing B1°F; gas, at 40- 
cm Hg. The effective counting sur- 
face was a horizontal slot 15-cm long X 
0.2-cem wide. As indicated in Fig. 1, 
the remaining area of the counter tube 
was covered with a shield that absorbed 
thermal neutrons and provided some 
collimation of the neutrons diffusing 
through the tank wall to the counter. 
Electrical pulses from the counter were 
fed into and recorded by a Tracerlab 
Amplisealer. This counter was used 
in all experiments except where other- 
wise noted. 
The other counter used was the 
scintillation counter depicted in Fig. 2. 
The detecting element was a “‘neutron 
phosphor” made by embedding silver- 
activated zine sulfide particles in a 
glyceryl-borate polyester containing 
naturally occurring boron (12). This 
counter will be described in detail in a 
future issue of Nuctronics. The 
neutron phosphor was placed horizon- 
tally against the face of a photomulti- 
plier tube. Photomultiplier tube out- 
put was fed into a cathode follower, 
156 
Counting-rate curves show effect of relative 
E 
5 
Fa 
a 
~~ 
2 
é 
10 
Liquid level by 
Intersection ye 
31.2 cm 31.3cm 
33.8 cm es 75cm 
Curve 
5. 
Counting Rate (cps) 
FIG. 4. Source is stationary 15-cm below liquid level as 
detector moves. 
Higher counting rates for B are caused by 
block of paraffin wax 6-cm thick covering source 
linear amplifier, discriminator and 
sealer. Although the active area of 
the scintillation counter was about half 
that of the boron-trifluoride counter, 
the scintillation counter was more 
efficient—it yielded more than double 
the count of the boron-trifluoride coun- 
ter under comparable conditions. 
Liquid-level measurements. An in- 
dependent liquid-level measurement 
was obtained in each experiment by a 
thermistor 1 mm in diameter. This 
was fixed to the counter, so that the 
thermistor moved vertically inside the 
tank wall as counter position was 
changed. Constant voltage was ap- 
plied across the thermistor. Thermis- 
tor current was relatively high in air 
and much lower when the thermistor 
was cooled by the water. Thus, as 
the thermistor was lowered towards 
the water, the position where the cur- 
rent suddenly decreased corresponded 
to the liquid level. The probable 
error in this measurement was esti- 
mated to be < + 1 mm. 
Counting-Rate Curves 
Curves of neutron counting rate vs 
counter position for various arrange- 
ments of source and counter are shown 
in Figs. 3-6. Counter position is 
plotted as the ordinate. The scale 
refers to an arbitrary zero that is very 
close to the bottom of the water tank. 
Liquid level, as determined by the 
thermistor, is shown as a horizontal 
line (the points below the line corre- 
spond to readings opposite the liquid). 
There is a definite discontinuity in 
the curve at the water level for most of 
the geometries studied. In the more 
favorable cases, where the curve is 
linear on either side of the sharp break, 
intersection of the two lines can be 
fixed to +1 mm or closer. In all these 
casé¢s the liquid level as indicated by 
the point of intersection on the curve 
agreed with the thermistor determina- 
tion to +1.5 mm or better. This 
agreement is within experimental error. 
Source below counter. Figure 3 
shows data obtained when the source 
is moved with the counter a fixed dis- 
tance below it. With the source very 
close to the counter (2.0-cm away, 
curve A), the counting rate increases 
rapidly as the counter approaches the 
liquid level from above, resulting in a 
smooth sigmoidal curve. This type of 
curve is an undesirable one, although 
the point of inflection appears to be 
close to the liquid level. The sharp 
rise in count as the water level is closely 
approached from above results from 
the fact that, when the source is just 
opposite the water level and is moving 
downward, the number of neutrons 
being confined by the water is increas- 
ing rapidly. When the source is placed 
somewhat farther below the counter 
(6.2-cm away, curve B), a discontinuity 
in the counting rate appears at the 
liquid level. With the source rela- 
tively far away (10.8 em, curve C), the 
counting rate increases very slowly as 
the counter approaches the liquid from 
above; therefore as the water level is 
