NUEN 605

Laboratory #7: Uranium Mass Measurement Using 3He Neutron Detectors

By
Ashraf Alfandi
Lab Partner: Joseph Sackman & Trevor Parker

Introduction

The purpose of this lab is to get introduced to different types of neutron detectors and see

how detector efficiency got affected by moderation. In addition, the lab was intended to measure

the enrichment of several unknown uranium samples using the comparator method. Moreover,

the mass of an unknown uranium sample was determined using the total neutron counting

method.

Materials and Methods

To begin, the detector was connected to the preamplifier (Pre-Amp), which was

connected the amplifier (Amp). The Amp was connected to the Single Channel Analyzer (SCA)

that was connected to the Timer/Counter. In order to check the pulse height, the oscilloscope (O-

Scope) was connected to the MCA. The testing arrangement was as seen in Figure 1. The serial

& model numbers are listed in Table 1.

Single
Pre- Timer/
Detector Amplifier Channel
Amplifier Counter
Analyzer

High
O-scope
Voltage

Figure 1.Testing configuration for 3He/BF3 detector measurement

1
 Table 1. Nuclear instrumentation modules and their serial numbers

Instrument Serial Number Model Number Manufacturer

Pre-Amp 13000932 2006 CANBERA
Oscilloscope C011816 TDS 1002C-EDU
SCA/Amp 1300094 2015B CANBERA
Timer/Counter 13161026 M996 ORTEC
3
He detector H3-05 DELTA
BF3 detector 2013-LND BF3-003 Delta AUG 60E/U

A 252Cf sample (metallic form) was used as a neutron source during the first half of the

experiment; see Table 2 for its radioactive properties. The source current activity, A(t), was

calculated using Eq. 1, taking into account the decay of the source,

A(t) = Ao et (Eq. 1)

ln(2)
= (Eq 2.)
T12

where,

Ao: The original source activity and equals (2.3 0.23) 106 neutron/second.

t: The decay time since 9th August 2005.

: The decay constant

T1/2: The source half-life.

Table 2. 252Cf source properties

Nuclide T1/2 (s) (s-1) Ao (n/s) Decay Time (s) A(t) (n/s)
252
Cf 8.34107 8.26106 (2.3 0.23) 106 3.85108 (8.86 0.88) 104

The power supply High Voltage (HV) was adjusted at 1150 V using the voltage dial. The

background counts were recorded by the 3He detector for 10 s, in which the Lower Level

Discriminator (LLD) level was gradually increased to eliminate background noise. The detector

2
efficiency was calculated for the 3He and BF3 detectors at different Polyethylene (PE)

thicknesses see Eq. 3. The PE slabs were placed in between and behind the detectors. The testing

configurations were as shown in Figure 2 and 3.

R
= YA (Eq. 3)

where,

R: the count rate recorded

Y: the source yield

Figure 2. Test configuration. The PE slabs were place between the detector and the neutron
source.

3
 Figure 3. Test configuration. The 3 PE slabs were placed behind the detector.

An optimum configuration was sat based on the detector efficiency, in which the PE slabs

were reconfigured to get the detector maximum efficiency (see Figures 4 & 5).

Figure 4. Moderation PE slabs Optimum configuration. One 2 PE slab were placed in all
directions except in front where two 2 inches were added.

4
 Figure 5. Moderation PE slabs optimum configuration. One more 2 PE slab was placed

on the top

For the second half of the experiment, natural enriched uranium metal rods were used for

measuring the sample mass, using the Total Neutron Counting (TNC) method. The testing

optimum configuration was adapting (see Figure 6), in which two 3He detectors were using. The

first sample rod counts recorded for 45 minutes, while 30 minutes for each of the other two rods.

Figure 6. TNC testing configuration. The first measurement using one uranium rod.

5
 The uranium mass was calculated according to Eq .4, as follows,

= Y (Eq. 4)

where,

m: the mass of the uranium sample (kg)

T: is the total number of the recorded counts rate (counts/s)

: is the detector efficiency

Y: is the source yield (n/kg.s)

ML: is the leakage multiplication factor and equals . 1.2 0.1

Results and Discussion

The 252Cf decays (3% of the time) by spontaneous fission, in which 3.7 neutrons

(average) are emitted. The produced fast neutrons have to be thermalized in order to be detected

by 3He and BF3 detectors. This is because 3He and 10B have high neutron absorption cross

section (a) in the thermal energy region (Eth ~ 0.035 eV). The counts were recorded for 10 s

using the CANBERA Counter/Timer, the data is shown in Table 3.

Table 3. 252Cf data at different PE thicknesses between the detector and the neutron source.

PE Thickness Number Count Rate, R Efficiency

R (s-1)
(in) of Counts -1
(s ) ()
0 1787 178.7 4.23 2.0210-3 2.0710-4
1 2570 257.0 5.07 2.9010-3 2.9610-4
2 2969 296.9 5.45 3.3510-3 3.4110-4
3 2478 247.8 4.98 2.8010-3 2.8610-4

The moderation thickness effect on the detection efficiency was investigated, by

calculating the detector efficiency at different moderation thicknesses (Eq. 3). The calculations

were performed using Excel software; the results are as shown in Table 3 and plotted as in

Figure 5.

6
 0.40%
0.35%
0.30%

Efficiency (%)
0.25%
0.20%
y = -1E-04x3 + 7E-05x2 + 0.0009x + 0.002
0.15%
R = 1
0.10%
0.05%
0.00%
0 0.5 1 1.5 2 2.5 3 3.5
PE Thickness (in)

Figure 7. Detector efficiency (%) as a function of PE thickness (in).

According to Figure 7, the detector efficiency increases with the thickness of

polyethylene, which means that the efficiency is increasing with moderation. However, the

efficiency values at 1 and 2 inches thick are statistically the same. This means that the efficiency

was not increasing upon PE addition. Moreover, the efficiency tends to decrease for larger

amount of moderation, because the neutron absorption increase, which decreases the number of

detected neutron and hence decrease detector efficiency.

The same measurement was conducted, where the PE slaps were placed behind the

detector as seen in Figure 3. The moderating material behind detector reflects the unscattered

neutrons, which increase the population of the neutrons near the detector, and hence increase the

detection efficiency (see Figure 8). It was noted that the detector efficiency was affected by the

thickness of moderation material. The 1 inch-thick material, was not enough to get the neutrons

thermalized, which result in lower detector efficiency. However, the detector efficiency

increased with moderation thickness.

7
 0.70%

0.60%

0.50%
Efficiency (%)
0.40%

0.30% y = -0.0004x3 + 0.0013x2 + 0.0006x + 0.002

R = 1

0.20%

0.10%

0.00%
0 0.5 1 1.5 2 2.5 3 3.5
PE Thickness (in)

Figure 8. Detector efficiency as a function of PE thickness (in).

According to the recordings, it was noted that the recorded number of counts was larger

comparing to the first measurement (see Table 4).

Table 4. 252Cf data at different polyethylene thickness behind the detector.

PE Thickness Number Count Rate, R Efficiency

R (s-1)
(in) of Counts (s-1) ()
1 3162 316.2 5.62 3.5710-3 3.6310-4
2 4837 483.7 6.95 5.4710-3 5.5210-4
3 4774 477.4 6.91 5.3910-3 5.4510-4

The 3He has higher absorption (a) than 10B, which reflects the higher efficiency of the
3
He detector over the BF3 detectors. Figure 9 shows that both of the two detectors shares the

same behavior of the moderation thickness effect on the efficiency. However, it should be

mentioned that the BF3 detectors are cheaper and hence more affordable. The 3He production is

very small compared to the demand, which justifies the high prices of 3He detectors.

8
 0.40%
y = -1E-04x3 + 7E-05x2 + 0.0009x + 0.002
0.35% R = 1

0.30%

0.25%
Effeciency (%)

0.20%
He Detector

0.15% BF3 Detector

0.10%
y = -7E-06x3 - 0.0003x2 + 0.001x + 0.0007
R = 1
0.05%

0.00%
0 0.5 1 1.5 2 2.5 3 3.5
PE Thickness (in)

Figure 9. The efficiency of the BF3 detector vs 3He detector at different PE thicknesses.
The new configuration (Figures 4 and 5) was meant to reflect as many neutron as it could

and minimizes the neutron leakage, which in turn would increase the neutron population near the

detector. It was found that the recorded number of counts increased by order of five to six, see

Table 5. Moreover, the detector efficiency increased to be 2.12% compared to 0.3 0.55% for

the previous configurations.

Table 5.252Cf data for optimum configuration.

Number Count Rate, R Efficiency

R (s-1)
of Counts (s-1) ()
18764 1876.4 23.73 2.1210-2 0.21410-2

For the TNC experiment, the background counts were recorded for 10 min, in which the
LLD was set at 0.30 V to eliminate the noise while the E adjusted to 10 V. The background
reading was 88 counts. Another 252Cf source (10Ci) was used for detector efficiency calculation
(Eq. 3) and was 2.9 0.1 %. The uranium mass was calculated according to the Eq. 4 and the
results as shown in Table 6

9
 Number Number of Recording time Count rate Uranium mass
of Rods Counts (min) (s-1) (kg)
1 6259 45 2.3 1.756
2 5868 30 3.26 2.45
3 6071 30 3.37 2.55

Conclusion

It was concluded that the number of slabs affects the detector efficiency, which indicated the

importance of the moderation in the neutron detector. An optimum configuration was decided

based on the detection efficiency in which the PE slabs were rearranged to get the maximum

efficiency. On the other hand, the mass of a uranium samples were measured, in which a natural

uranium samples were recorded for 45 min, 30 min, and 30 min, consecutively. The uranium

mass were found different from the actual one (1.8 kg/rod) due to the recording time. The

Uranium metal neutron yield is relatively small which require long time recording to get the

accurate value. Moreover, the errors associated with taking measurement has to be considered in

the results evaluation.

10
11