To verify the self-indication method, we have performed three kinds of experiments using a 46 MeV electron linear accelerator (linac) at the Kyoto University Research Reactor Institute. The experimental arrangement is shown in Fig. 3.1. Pulsed neutrons were produced from a water-cooled photo-neutron target assembly, 5 cm in diameter and 6 cm long, which was composed of 12 sheets of tantalum plates with total thickness of 29 mm . This target was set at the center of an octagonal water tank, 30 cm long and 10 cm thick, as a neutron moderator. The linac was operated with a repetition rate of 50 Hz, a pulse width of 100 ns, a peak current of 5 A, and an electron energy of about 30 MeV. We used a flight path in the direction of 135 o to the linac electron beam. To reduce the gamma flash generated
Fig. 3.1 Experimental arrangement for the time-of-flight (TOF) measurement
by the electron burst from the target, a lead block, 7 cm in diameter and 10 cm long, was set in front of the entrance of the flight tube.
First, a gold foil 10, 20, 30, 40, or 50 μm thick was located as a sample at a flight length of 11.0 m, where the neutron beam was well collimated to a diameter of 24 mm. An indicator located at a flight length of 12.7 m was surrounded by a Bi4Ge3O12 (BGO) assembly, which consists of 12 scintillation bricks each 5 x 5 x 7.5 cm3 . Prompt-capture γ-rays from the indicator were detected with the BGO assembly in the TOF measurement. A 10B plug 8 mm thick or gold foil 50 μm thick was used as an indicator. Because the former thick indicator can absorb most neutrons with energies below the epi-thermal region, it was equivalent to the conventional NRTA. In the latter case, it was the self-indication measurement. The area densities of the samples with different thickness were estimated by area analysis for the 4.9 eV resonance of 197Au.
As the next step, a 50-μm-thick silver foil was added to a 10-μm-thick gold foil to form a sample and the area density of the gold foil was measured. It is worth noting that silver has a large resonance at 5.2 eV, close to the 4.9 eV resonance of 197Au. The 10B plug 8 mm thick or a gold foil 50 μm thick was also used as an indicator. Moreover, we demonstrated a nondestructive assay for nuclear materials using a mixture composed of a natural uranium plate and sealed minor actinide samples of 237Np and 243Am. The natural uranium plate was 10 x 20 mm2 and weighed 5.8 g. The samples of 237Np and 243Am were oxide powder, which was pressed into a pellet 20 mm in diameter and encapsulated in an aluminum diskshaped container 30 mm in diameter with 0.5-mm-thick walls. The activities of 237Np and 243Am were 26 and 868 MBq, respectively. In the third measurement, the 10B plug 8 mm thick or a natural uranium plate of 10 x 20 mm2 and weight 5.8 g was used as an indicator.
Results and Discussion
The TOF spectra obtained by both methods, NRTA (dotted lines) and the selfindication method (solid lines), around the 4.9 eV resonance of 197Au are shown in Fig. 3.2. Neutron absorption resulting from the 4.9 eV resonance of 197Au can be observed as a dip and a lack of peak for the NRTA and the self-indication method, respectively. Here the net area ratio R, which is defined as the ratio of resonance absorption to the number of incident neutrons, is defined by
where Ci is the net counting rate of the ith channel in the TOF measurement and the subscripts “in” and “out” mean “with” and “without” sample. Background events were estimated by TOF measurement without an indicator and subtracted from the foreground TOF spectrum. The 4.9 eV resonance peak ranges from the Iminth to Imaxth channel. The net area ratios for the NRTA and the self-indication method can be expressed as follows:
Fig. 3.2 Comparison of TOF spectra obtained by transmission method and self-indication method for 4.9 eV resonance of 197Au
Fig. 3.3 Relationship between net area ratio and sample thickness
where n and nind denote the thickness (the number of target nuclide per unit area) of the sample and indicator, respectively. The quantities σtot and σcap represent the energy-dependent neutron total and capture cross sections of the target nuclide, respectively. The integration is performed over the resonance peak region. By applying Eqs. (3.2) and (3.3), the relationships between the net area ratio and the thickness were obtained using point-wise cross-section data of JENDL-4.0 , as shown in Fig. 3.3. If nσtot is not large, the net area ratio is proportional to the sample thickness. However, it converges to unity and loses information about thickness as nσtot becomes larger. The thickness of each gold foil sample was derived from the relationship of Fig. 3.3 and the value of R was determined by experiment. Figure 3.4 shows the results of quantitative examination. It was confirmed that the thickness of the target nuclide can be determined by both methods within 3 % accuracy. The accuracy can be improved further by using a smaller resonance (nσtot is not large).
The TOF spectra with silver and gold samples are shown with the NRTA method in Fig. 3.5 and with the self-indication method in Fig. 3.6. In NRTA, the dips of the 4.9 eV resonance of 197Au and the 5.2 eV resonance of 109Ag overlapped around 400 ch. in Fig. 3.5. In the self-indication method, the contribution from impurity was suppressed and a weak 58 eV resonance of 197Au was emphasized around 120 ch. (Fig. 3.6). The TOF spectra for the mixture composed of natU, 237Np, and 243Am are shown in Figs. 3.7 and 3.8. Although many resonance dips caused by impurities of 237Np and 243Am were observed (Fig. 3.7), there are no differences
Fig. 3.5 TOF spectra obtained with neutron resonance transmission analysis (NRTA) for pure gold and mixture of gold and silver
between TOF spectra with only natU (blue line) and the mixture (red line) (Fig. 3.8). This result indicates that the self-indication TOF spectrum was not greatly influenced by nuclide impurity. It was experimentally shown that the contribution from the other nuclide can be remarkably suppressed by applying the selfindication method.
Fig. 3.6 TOF spectra obtained with the self-indication method for pure gold and mixture of gold and silver
Fig. 3.7 TOF spectra obtained with NRTA for natU and mixture of natU, 237Np, and 243Am
Fig. 3.8 TOF spectra obtained with the self-indication method for natU and mixture of natU, 237Np, and 243Am
In this work, we proposed a new concept of the “self-indication method” as a complementary nondestructive assay for the fuel debris of Fukushima Daiichi NPP. We carried out experimental validation for application of the self-indication method. It was confirmed that the area density (thickness of the target nuclide) can be determined within 3 % accuracy by simple area analysis without a resonance fitting process. Moreover, it was experimentally shown that the contribution from the other nuclide can be remarkably suppressed by applying the self-indication method. The self-indication method combined with the TOF technique will be a useful tool for nondestructive assaying of the distribution of nuclear material in the melted fuel debris, which contains many impurities and has high activities.