Measurements and Calculations of Neutron Energy Spectra Behind Polyethylene Shields Bombarded by 40- and 65-MeV Quasi-Monoenergetic Neutron Sources

Transcription

1 Journal of NUCLEAR SCIENCE and TECHNOLOGY, Vol. 34, No. 4, p (April 1997) Measurements and Calculations of Neutron Energy Spectra Behind Polyethylene Shields Bombarded by 40- and 65-MeV Quasi-Monoenergetic Neutron Sources Noriaki NAKAO*1,t, Makoto NAKAO*2, Hiroshi NAKASHIMA*3, Susumu TANAKA*4, Yukio SAKAMOTO*3, Yoshihiro NAKANE*3, Shun-ichi TANAKA*3 and Takashi NAKAMURA*2 *1 Institute for Nuclear Study, University of Tokyo *2 Cyclotron and Radioisotope Center, Tohoku University *3 Tokai Establishment, Japan Atomic Energy Research Institute *4 Takasaki Establishment, Japan Atomic Energy Research Institute (Received September 25, 1996) Measurements of neutron energy spectra behind 30.5-, 61.0-, , cm-thick polyethylene shields bombarded by 40- and 65-MeV quasi-monoenergetic neutrons are performed at the 90-MeV AVF cyclotron of the TIARA (Takasaki Ion Accelerator for Advanced Radiation Application) at JAERI (Japan Atomic Energy Research Institute). Source neutrons are produced at 3.6- and 5.2-mm-thick 7Li targets bombarded by 43- and 68-MeV protons, respectively. A BC501A organic liquid scintillator and multi-moderator spectrometer with a 3He counter (Bonner ball) are used for spectrometry of transmitted neutrons and their energy spectra are obtained with the unfolding technique. The energy spectra from a few MeV up to a peak energy are obtained by the BC501A scintillator measurement and those below a few MeV down to thermal energy are obtained by the Bonner ball measurement. The measurements are performed on the neutron beam axis and at off-center positions, and attenuation profiles of neutron fluxes along the beam axis are obtained. The MORSE Monte Carlo calculations are performed with the DLC119/HILO86 multi-group cross section library for comparison with the measured data. The calculation generally gives a little overestimated fluxes, and a few % longer attenuation lengths of peak flux and dose equivalent. KEYWORDS: neutron energy spectra, shielding materials, polyethylene, shields, quasimonoenergetic neutrons, organic liquid scintillator, Bonner ball, transmitted neutrons, MORSE Monte Carlo, MORSE potential, DLC119/HILO86, dose equivalent, MeV range , experimental data, cross sections, caparative evaluations I. INTRODUCTION As particle accelerators become capable of producing beams of higher energy and intensity, the radiation shielding becomes even more important for protection of the people who work around or live near an accelerator. Utilization of accelerators is nowadays expanding to a variety of applications and some big projects of high energy hadron accelerator construction have started in our country. More accurate shielding design for higher energy particle accelerator based on a sophisticated calculation code and accurate nuclear data are therefore required to save the construction costs. Experimental shielding benchmark data are very important to improve the calculational model and to revise *1Midori -cho, Tanashi-shi 188. *2Aramaki, Aoba-ku, Sendai 980. *3Tokai -mura, Naka-gun, Ibaraki-ken *4 Watanuki -cho, Takasaki-shi Corresponding author, Tel. t , Fax , nakaon@ins.u-tokyo.ac.jp nuclear data for shielding calculations of high energy neutrons in accelerator facilities. Some shielding experiments have ever been performed by Shin et al.(1) and Uwamino et al.(2) using white spectral neutrons whose maximum energies were -50MeV, and by Ishikawa et al.(3) using 22- and 32-MeV quasi-monoenergetic neutrons for concrete and iron shields. Experimental shielding data of quasi-monoenergetic neutrons of higher energy are, however, still required for checking the accuracy of calculation codes and data libraries. We have already performed the shielding experiments, for concrete(4) and iron(5) by using 40- and 65-MeV quasi-monoenergetic neutron sources developed at the 90-MeV AVF cyclotron TIARA facility at Japan Atomic Energy Research Institute (JAERI) as the Universities-JAERI co-operative project research programme. These unique neutron fields were specially designed for shielding and cross-section experiments and the monoenergetic peak fluxes and the energy spectra of neutron sources produced by the 7Li(p, n) reaction have well been estimated with var- 348

2 Measurements and Calculations of Neutron Energy Spectra Behind Polyethylene Shields 349 ious repeated experiments. In this study following the concrete and iron shielding experiments, we measured the energy spectra of neutrons behind polyethylene shields up to 183.0cm thickness with 40- and 65- MeV quasi-monoenergetic neutron sources at the AVF cyclotron TIARA facility at JAERI. Iron shield is good for high energy neutron attenuation because of large inelastic cross section and high density, but brings out a dominant peak of 10-keV- to 1-MeV-energy in the neutron spectrum behind it. Polyethylene shield is often used behind the iron shield to reduce these low energy neutrons, because of large content of hydrogen which attenuates them most effectively. We also provide the results calculated by using the MORSE-CG Monte Carlo code to compare with our experimental data. These experimental data can also be used as benchmark data to investigate the accuracy of calculation codes and the cross-section data libraries. II. EXPERIMENT The shielding experiment was performed at the neutron beam course named the LC-course of the 90-MeV AVF cyclotron facility, TIARA at JAERI. The crosssectional view of the beam course is shown in Fig. 1. Accelerated protons were transported to the neutron beam course, and bombarded thin 7Li targets (99.9% enriched) equipped at the target chamber in the accelerator room. The protons that penetrated the target were bent down toward the Faraday cup in the beam dump by the clearing magnet, and their integrated charges were measured with the current integrator. The neutrons produced at 0 deg. in the 7Li target were transported to the shielding experimental space at the 3rd light-ion-room through the 10.9-cm-diameter iron rotary shutter collimator inserted in the 2.2-m-thick concrete shielding wall. In this experiment, we used quasi-monoenergetic neutron sources generated by protons of 43 and 68MeV energies bombarding 3.6- and 5.2-mm-thick targets, respectively, with 2-MeV energy loss of primary protons. The energy spectra and peak fluxes of the source neutrons have already been measured and reported in Refs.(4) and (5). Figure 2 gives the energy spectra of source neutrons obtained by the TOF measurement with the BC501A organic liquid scintillator, and Table 1 gives the absolute peak fluxes estimated by the measurements with PRT (Proton Recoil counter Telescope)(6). As shown in Fig. 1, two fission chambers of 238U and 232Th were equipped near the target and were calibrated to the integrated proton beam current, and they were used as neutron monitors during the experiment. They have thresholds at about 1MeV and comparably large and flat cross sections in several tens MeV energy range. Count rates of two fission chambers are proportional to the source neutron flux and proton beam current which cannot be measured well in low intensity because of the dark current. The calibrated values during this experiment are also given in Table 1. Using these values, the fluences of source neutrons injected onto the shield assembly were obtained from the monitor counts. The cm-widex118.0-cm-highx30.5-cm-thick Fig. 1 Cross-sectional view of the neutron beam course at TIARA facility Fig. 2 Energy spectra of quasi-monoenergetic neutron sources measured with the BC501A detector using the TOF method The spectra below 7MeV were extrapolated on an assumption that energy spectrum is flat. Table 1 Peak fluxes of source neutrons measured with PRT detector and calibrated reaction rates of 238U fission chamber t Neutron flux at 403cm from the target VOL. 34, NO. 4, APRIL 1997

3 350 N. NAKAO et al. polyethylene slabs were assembled to the to cm thick shields. The density and the atomic composition are tabulated in Table 2. The polyethylene shields were fixed on a movable stand shown in Fig. 1, and were inserted into 120-cm-widex120-cm-highx120-cmdeep experimental hole to sit in contact with the neutron beam exit. The top views of the experimental arrangement are shown in Fig. 3. Figure 3(a) exemplifies the measurements on the beam axis for cm-thick shield. In thinner shield experiments, the outer slabs were removed and the neutron detectors were placed just Table 2 Composition and density of the polyethylene shield behind the shield assembly on the beam axis. On the other hand, as shown in Fig. 3(b), the measurements at the distances off the beam axis for and 61.0-cm thick shields were performed with using the additional iron collimators to measure the neutrons scattered at a large angle. The additional collimators were needed to reduce the contribution of neutrons leaking through the iron ball and iron sand fillers, and small gaps between the shields in rotary shutter of iron and polyethylene (see Fig. 1). As shown in the Figs. 3(a) and 3(b), narrow gaps of 0.5 to 1.6cm widths were settled between the polyethylene slabs for inserting small detectors used by other experimental group. Measurements of neutron energy spectra behind the polyethylene shields were performed with an organic liquid scintillator and a multi-moderator spectrometer. A 12.7-cm-diameter by 12.7-cm-long cyrindrical organic liquid scintillator, BC501A (Bicron Co., Ltd.), coupled Fig. 3 Experimental arrangement of shield assembly and detector (a) on measurement on the beam axis and (b) on measurement at distance off the beam axis using additional iron collimator (43MeV: 40cm thick, 68MeV: 80cm thick) JOURNAL OF NUCLEAR SCIENCE AND TECHNOLOGY

4 Measurements and Calculations of Neutron Energy Spectra Behind Polyethylene Shields 351 Fig. 4 Block diagram of electronics for the BC501A scintillator measurement with an R4144 photomultiplier (Hamamatsu Photonics Co., Ltd.) was placed just behind the shield surface, as shown in Fig. 3, for neutron spectrometry of energy above a few MeV. Two-dimensional data of pulse height and rise-time of the scintillation light outputs were taken in an event-by-event mode by the ND9900MPA 8/16 (CAMBERRA Co., Ltd.) coupled with the VAX Station 3100 (DEC Co., Ltd.) operating system. A block diagram of measuring circuit is shown in Fig. 4. The rise-time information was utilized to the n-g discrimination on the off-line analysis. The pulse height threshold for taking event data was determined by an SCA module, and the pulses above the threshold were counted also by the scaler for the correction of event data storing loss. A multi-moderator spectrometer (Bonner ball) was also used to measure the neutron energy spectra down to thermal energy. The spectrometer consists of a 5.08-cmdiameter spherical 3He proportional counter (10atm) made by LND Inc. surrounded with four spherical polyethylene moderators of 1.5-, 3.0-, 5.0- and 9.0-cm thicknesses and a bare counter, from which the five count rates above the g-ray discrimination level for each moderator thickness were measured only on the beam axis behind the shield assembly. The detector position and geometrical condition in each measurement are tabulated in Table 3. III. DATA ANALYSIS The pulse height distributions of the BC501A detector induced by neutrons were obtained by eliminating the -ray events. The energy calibration of the pulse height g distribution was performed using the Compton edges of 1.24 and 4.43MeV g-rays from 22Na and 241Am-Be sources, respectively, and the recoiled proton edges of 40 and 65MeV neutron sources. Then the neutron energy spectra were obtained from these pulse-height distributions using the FERDOU(7) unfolding code and the measured response matrix(8). From the five count rates of Bonner ball, the neutron energy spectra were obtained using the SAND-2(9) unfolding code and the cited response funcitons(10). The initial guess spectra for this unfolding were obtained from the Monte Carlo calculations described in the next section. The neutron spectra and the count rates were finally normalized to the integrated beam currents in the unit of micro Coulomb which were converted from the flux monitor counts of each experiment using the calibrated values in Table 1. Table 3 Experimental geometry conditions of shield thickness, collimator thickness and detector positions VOL. 34, NO. 4, APRIL 1997

5 352 N. NAKAO et al. IV. CALCULATIONS For comparison with our measured neutron energy spectra, Monte Carlo calculations were performed with the MORSE-CG code(11) and the DLC119/HILO86(12) multigroup cross-section library. The DLC119 has an energy structure of 66 groups between thermal and 400- MeV energies for neutrons and of 22 groups up to 20 MeV for g-rays and is of a P5 Legendre expansion. The macroscopic cross-section data of polyethylene used in the calculations was made from the DLC119 library using the atomic composition given in Table 2. The HILO86R(13) library, which is a macroscopic group constants copied from DLC119 for En>19.6MeV and collapsed from JSSTDL(14) for En<19.6MeV with considering the self-shielding factor, was also utilized for the cross-section data of the additional iron collimator. As the energy spectra of source neutrons were only measured above 7MeV by the TOF method, those below 7MeV down to thermal energy were approximated to be constant to the values at 7MeV (see Fig. 2) and thus-extrapolated source spectra were used for the calculations. The calculational geometry is shown in Fig. 5. The source neutrons were assumed to be emitted only in a very sharp cone of 5.94x10-4sr along the beam line as seen in Fig. 5(a), which means the neutron beam spread of 10.9cm diameter at the beam exit, 396cm from the source point. Considering good statistics and the voluminous detector, we placed the next event track length estimator at each detector position for flux estimations cmdiameter by 12.7-cm-long cyrindrical estimators, corresponding to the actual BC501A detector used in the experiment, were placed on the beam axis and at 20- and 40-cm off the beam axis on the calculations with the additional collimator, as shown in Fig. 5(b). All the results were normalized to the proton beam charges in the unit of micro Coulomb for comparison with the measured results, using the values of the peak fluxes of source neutrons at the 403cm from the 7Li target given in Table 1. The calculations were performed by use of the SUN Ultra-1 computer and the conditions of history number, FSD (fractional standard deviation) for total fluxes and computing time are tabulated in Table 4. In order to save the computing time, the calculations of energy spectra higher than thermal energy group were performed except the calculation for 30.5-cm-thick shield. The calculations only for the thermal energy group were separately performed with 1/10 histories of the values in Table 4 and the results were connected to each spectrum above thermal energy. V. RESULTS AND DISCUSSIONS 1. Transmitted Neutron Spectra on the Beam Axis The measured and calculated neutron energy spectra on the beam axis transmitted through polyethylene shields of various thicknesses are shown in Figs. 6 and 7 for 43- and 68-MeV p-li neutron sources, respec- Fig. 5 Calculational geometry on the MORSE Monte Carlo calculation (a) Source particles were emitted within a solid angle on 5.94x10-4sr. Tc is a thickness of additional collimator which are given in Table 3. On the calculation without additional collimator, Tc is 0 and the surface of the polyethylene shield is at the 403cm. The gaps between shields are also considered. (b) Cyrindrical flux estimators were placed behind the shield surface on and off beam axis. JOURNAL OF NUCLEAR SCIENCE AND TECHNOLOGY

6 Measurements and Calculations of Neutron Energy Spectra Behind Polyethylene Shields 353 Table 4 Calculational conditions of history, FSD for total fluxes and computing time t : Calculation down to thermal group, x: Calculation except thermal group Fig. 6 Comparison of measured and calculated neutron energy spectra behind various thick shield using 43MeV p-li neutrons (a) The energy spectra measured with Bonner ball detector are shown in the energy range from thermal to peak energy. (b) The energy spectra measured with BC501A detector are shown in the energy range above a few MeV. tively. The Bonner ball detector gives the spectra over the entire energy range from thermal to peak energy in Figs. 6(a) and 7(a), and the BC501A detector gives the spectra above a few MeV in Figs. 6(b) and 7(b). The energy spectra only below 10MeV given by the Bonner ball were used for comparison because of the low sensitivity of the Bonner ball in higher energy. Figures 6 and 7 indicate the general tendency that the MORSE calculation gives rather good agreement with the measured results. By looking in detail, the calculated spectra are in good agreement with the BC501A spectra above a few MeV for thinner polyethylene shield but become larger than the BC501A spectra with the shield thickness. Comparisons of measured and calculated neutron fluxes integrated in three energy regions; the peak, the continuum and the low energy regions, are given in Table 5 and are shown in Figs. 8(a)-(c). In the peak and continuum regions, the MORSE calculations were compared with the BC501A results and in the low energy region with the Bonner ball results. The experimental error of the BC501A energy spectra includes only the errors of coming from the spectrum unfolding process, VOL. 34, NO. 4, APRIL 1997

7 354 N. NAKAO et al. Fig. 7 Comparison of measured and calculated neutron energy spectra behind various thick shield using 68MeV p-li neutrons (a) The energy spectra measured with Bonner ball detector are shown in the energy range from thermal to peak energy. (b) The energy spectra measured with BC501A detector are shown in energy range above a few MeV. and the counting statistics in the measurement. The error of the absolute source neutron intensity measured by PRT (3 or 5%, see Table 1), the error of the conversion factor of fluence monitor count to the proton beam charge (3%), the error of the neutron penetration factor through objects on the beam line (3%), that is the small detectors of the other group, and the statistical error of the fluence monitor count (<1%) were neglected. On the experimental errors of neutron fluxes integrated in peak and continuum regions (see Table 5), however, all of these errors were considered. The experimental errors of Bonner ball could not be obtained because the SAND- 2 unfolding code does not give them. The errors of the calculated values were given from the statistical errors in the Monte Carlo calculation. In Table 5 and Fig. 8, the integrated fluxes in the three energy regions show about 10% to a factor of 2 discrepancy between calculation and experiment, and the calculations give the larger values than the experiments. The descrepancy becomes larger with increasing the polyethylene thickness. Experimental and calculated dose equivalents were also estimated by folding the neutron energy spectra with the neutron-flux-to-dose-equivalent conversion factor cited from ICRP21 publication(15), and their comparisons are tabulated in Table 6 and are shown in Fig. 9. In the experimantal dose estimation, the measured energy spectra with the BC501A were used above 10 MeV and those with the Bonner ball were employed below 10MeV. From Table 6, it is clarified that neutrons above 10MeV which occupy around 90% of total flux have dominant contribution to dose. The ratios of calculated to experimental values are therefore similar to those of peak neutron fluxes, and are 1.07 to The experimental errors of dose-equivalent include only those of BC501A in which all errors refered above are considered. Attenuation profiles only for peak neutron flux and dose-equivalent, as shown in Figs. 8(a) and 9, could be fitted to single exponential curves, and the obtained attenuation lengths of polyethylene shield are given in Table 7 for 40- and 65-MeV neutrons. The calculated attenuation lengths are 7% and 3% longer than the experimental ones for 40- and 65-MeV peak neutron flux, respectively, and are 7% and 4% longer for dose equivalent, respectively. This clarified that the calculation gives the safer side estimation of the dose equivalent outside the polyethylene shield. Measured and calculated count rates of the 3He counter with and without four kinds of moderators are compared in Figs. 10(a),(b) for 43- and 68-MeV p-li JOURNAL OF NUCLEAR SCIENCE AND TECHNOLOGY

8 Measurements and Calculations of Neutron Energy Spectra Behind Polyethylene Shields 355 Table 5 Comparisons of experimental and calculated neutron fluxes behind the polyethylene shield integrated in various energy ranges Peak fluxes at 0cm thickness are source neutron fluxes at 403cm from the target. Experimental fluxes at peak and continuum region are estimated with the results of BC501A scintillator and those at low energy region are estimated with the results of Bonner ball detector. Experimental errors at low energy region are not given by the SAND-2 code. t Read as 2.12x104 Table 6 Comparisons of dose-equivalent behind the polyethylene shield estimated with calculated and measured energy spectra Measured dose equivalents were estimated with using the neutron energy spectra by Bonner ball (En<10MeV) and BC501A scintillator (10MeV<En). The values at 0cm thickness were estimated with the energy spectra of source neutrons at 403cm from the target. t Read as 1.34x10-1 VOL. 34, NO. 4, APRIL 1997

9 356 N. NAKAO et al. Fig. 9 Comparison of experimental and calculated dose-equivalent The fitted curves are also shown. Table 7 Comparisons of experimental and calculated attenuation lengths of peak neutron flux and dose equivalent for the polyethylene shield neutrons. The calculated count rates were obtained by folding the calculated energy spectra multiplied by the neutron response functions of the Bonner ball. The statistical flax errors given by the MORSE calculation were considered to estimate the errors of count rate. From the comparison of count rate of the 9-cm-thick moderator detector, it is considered that calculations overestimate a factor of 2 in the energy range of eV, where the 9-cm-thick moderator has the dominant sensitivity. From Figs. 8, 9 and 10, it can be found that the slopes of the calculated attenuation curves are gentler than those of measured attenuation curves all for peak flux, doseequivalent and count rate. Fig. 8 Comparison of experimental and calculated fluxes integrated in the energy range of (a) peak region, (b) continuum region and (c) low energy region. Fitted curves of peak flux are also shown in (a). 2. Transmitted Neutron Spectra off the Beam Axis For the experiments of and 61.0cm thick polyethylene shields with the additional iron collimator, the neutron energy spectra measured with the BC501A detector on the shield surface on and off the beam axis are compared with the MORSE calculation in Table 8 and in Figs. 11 and 12 for 43 - and 68-MeV p-li neutrons, respectively. The calculations tend to give underestimated peak fluxes behind 30.5cm thick polyethylene shield off the beam axis for 43-MeV although they give slightly overestimated fluxes on the JOURNAL OF NUCLEAR SCIENCE AND TECHNOLOGY

10 Measurements and Calculations of Neutron Energy Spectra Behind Polyethylene Shields 357 Fig. 10 Comparison of measured and calculated count rates of 3He counter with four kinds of moderators and a bare counter for neutrons behind the various thick shield using (a) 43 MeV p-li neutrons and (b) 68MeV p-li neutrons Table 8 Comparisons of experimental and calculated fluxes on and off the beam axis integrated in the peak and the continuum regions on the geometry using additional iron collimator Experimental fluxes were estimated with the results of the BC501A detector. t Read as 3.16x103 VOL. 34, NO. 4, APRIL 1997

11 358 N. NAKAO et al. Fig. 11 Comparison of measured and calculated neutron energy spectra on the shield surface on and off the beam axis behind (a) 30.5-cm-thick and (b) 61.0-cm-thick shield using 43MeV p-li neutrons The energy spectra measured with BC501A detector are shown in the energy range above a few MeV. beam axis. As a whole, the MORSE calculation gives rather good agreement with the experiment, but the overestimation of the calculation results become larger for 61.0cm thick polyethylene for 68MeV p-li neutrons (Fig. 12(b)). This facts lead to a conclusion that angular distributions given by the Legendre expansion coefficients in the calculation have some problems. VI. CONCLUSION Neutron energy spectra behind polyethylene shields of 30.5-, 61.0-, , cm thickness were measured with a BC501A scintillator and a Bonner ball detector using quasi-monoenergetic neutron sources generated by 43- and 68-MeV protons bombarding thin 7Li targets at the TIARA facility. Neutron energy spectra in the energy range from thermal energy to peak neutron energy were obtained by these two detectors using the unfolding method. From the measured data, attenuation profiles of neutron fluxes in three energy regions, dose equivalent and count rates of the 3He counter with and without four kinds of moderater were obtained. This experiment provides the absolute values of neutron energy spectra and count rates of the Bonner ball behind polyethylene shields and they will be useful as an integral benchmark data. Neutron transport calculations were performed by using the MORSE-CG Monte Carlo code and the DLC119/HILO86 multigroup cross-section library. The calculated energy spectra behind the polyethylene shield agreed in general with the measured results and the discrepancy between experimental and calculated fluxes were from a few % up to a factor of -2. ACKNOWLEDGMENT We would like to express our appreciation to Dr. Tomohiko Iwasaki, Mr. Yoshimasa Sakuya and Mr. Yasushi Nauchi of Tohoku University for their great help during the experiment. We would like to give our appreciation to Mr. Shin-ichiro Meigo and Mr. Hiroshi Takada of Japan Atomic Energy Research Institute (JAERI) for their helpful advices for the measurement and the analysis. We wish to thank the cyclotron operators at the TIARA facility for their helpful co-operation and technical advices. This work has been supported by the Universities-JAERI Collaborative Project Research Program. JOURNAL OF NUCLEAR SCIENCE AND TECHNOLOGY

12 Measurements and Calculations of Neutron Energy Spectra Behind Polyethylene Shields 359 Fig. 12 Comparison of measured and calculated neutron energy spectra on the shield surface on and off the beam axis behind (a) 30.5-cm-thick and (b) 61.0-cm-thick shield using 68MeV p-li neutrons The energy spectra measured with BC501A detector are shown in the energy range above a few MeV. -REFERENCES- (1) Shin, K., Ishii, Y., Uwamino, Y., Sakai, H., Numata, S.: Radiat. Prot. Dosim., 37, 3, 175 (1991). (2) Uwamino, Y., Nakamura, T., Shin, K.: Nucl. Sci. Eng., 80, 360 (1982). (3) Ishikawa, T., Miyama, Y., Nakamura, T.: Nucl. Sci. Eng., 16, 278 (1994). (4) Nakao, N., Nakashima, H., Nakamura, T., Tanaka, Sh., Tanaka, Su., Shin, K., Baba, M., Sakamoto, Y., Nakane, Y.: Nucl. Sci. Eng., 124, 228 (1996). (5) Nakashima, H., Nakao, N., Tanaka, Sh., Nakamura, T., Shin, K., Tanaka, Su., Sakamoto, Y., Takada, H., Meigo, Y., Nakane, Y., Baba, M.: Nucl. Sci. Eng., 124, 243 (1996). (6) Baba, M., Iwasaki, T., Kiyosumi, T., Yoshioka, Y., Matsuyama, S., Hirakawa, N., Nakamura, T., Tanaka, Su., Tanaka, Sh., Nakashima, H., Meigo, S., Tanaka, R.: Proc. Int. Conf. Nuclear Data for Science and Technology, Gatlinburg, Tennessee, May 9-13, 1994, Val.1, p.90 (1994). (7) Shin, K., Uwamino, Y., Hyodo, T.: Nucl. Technol., 53, 78 (1981). (8) Nakao, N., Nakamura, T., Baba, M., Uwamino, Y., Nakanishi, N., Nakashima, H., Tanaka, Sh.: Nucl. Instrum. Methods., A362, 454 (1995). (9) McElroy, W.N., Berg, S., Crocktt, T., Hawkins, R.G.: AFWL-TR-67-41, Vol.1-4, (1967). (10) Uwamino, Y., Nakamura, T., Hara, A.: Nucl. Instrum. Methods., A239, 299 (1985). (11) Straker, G.R., Stevens, P.N., Irving, D.C., Cain, V.R.: ORNL-4585, (1970). (12) Alsmiller, Jr. R.G., Barns, J.M., Drischler, J.D.: ORNL/TM-9801, (1986). (13) Kotegawa, H., Nakane, Y., Hasegawa, A., Tanaka, Sh.: JAERI-M , (1993). (14) Hasegawa, A.: Nucl. Data Sci. Technol., p.232 (1991). (15) "Data for Protection Against Ionizing Radiation from External Sources: Supplement to ICRP Publication 15", Publ. 21, ICRP, (1971). VOL. 34, NO. 4, APRIL 1997