Measurement of Neutron-Production Double-Differential Cross Sections for Nuclear Spallation Reaction Induced by 0.8, 1.5 and 3.

Transcription

1 Journal of NUCLEAR SCIENCE and TECHNOLOGY, Vol. 34, No. 6, p (June 1997) Measurement of Neutron-Production Double-Differential Cross Sections for Nuclear Spallation Reaction Induced by 0.8, 1.5 and 3.0 GeV Protons Kenji ISHIBASHI*1,t1, Hiroshi TAKADA*2, Tatsushi NAKAMOTO*1,t2, Nobuhiro SHIGYO*1, Keisuke MAEHATA*1, Naruhiro MATSUFUJI*1,t3, Shin-ichiro MEIGO*2, Satoshi CHIBA*2, Masaharu NUMAJIRI*3, Yukinobu WATANABE*4 and Takashi NAKAMURA*5 *1Department of Nuclear Engineering, Kyushu University *2 Japan Atomic Energy Research Institute *3 National Laboratory for High Energy Physics *4 Energy Conversion Engineering, Kyushu University *5 Cyclotron Radioisotope Center, Tohoku University (Received November 15, 1996) Neutron-production double-differential cross sections were measured for the spallation reaction induced by 0.8, 1.5 and 3.0GeV protons on C, Al, Fe, In and Pb targets. The experiments were performed by time-of-flight technique with a typical flight path length of 1m, and the cross sections were obtained with energy resolutions better than 8% at neutron energies below 100MeV. The experimental data were compared with the results of calculation codes based on an intranuclear-cascade-evaporation model. Adoption of the in-medium effect on nucleon-nucleon cross section in the cascade calculation improves the agreement between the calculated results and experimental data particularly in the case of 0.8-GeV proton incidence. For protons above 1GeV, however, the calculations typically twice overestimate the cross sections in the emitted neutron energy region of 10 to 30MeV. KEYWORDS: neutron-production double-differential cross sections, incident proton, NE213, intranuclear-cascade-evaporation model, GeV range 01-10, MeV range , experimental data, neutron energy, carbon target, aluminium target, iron target, indium target, lead target, time-of-flight method, spallation, evaluations I. INTRODUCTION A great number of spallation neutrons are emitted from a target nucleus by proton bombardment with incident energies around GeV. Studies on the spallation reaction have recently been made for such applications as spallation neutron sources and accelerator-driven transmutation systems. Nuclear data are required for these purposes in the energy region up to a few GeV. The nuclear data libraries such as ENDF/B-VI(1) and JENDL-3.1(2) have been evaluated up to 20MeV, but *1 Hak ozaki, Higashi-ku, Fukuoka *2 Tok ai-mura, Naka-gun, Ibaraki-ken *3 Oho, Tsukuba-shi 305. Kasuga-koen, Kasuga-shi * *5Ar amaki, Aoba-ku, Sendai Corresponding author, Tel. t , F ax , kisibasi@kune2a.nucl.kyushu-u.ac.jp t2 Present address: National Laboratory for High Energy Physics, Oho, Tsukuba-shi Pres t ent address: National Institute of Radiological Sciences, Anakawa, Inage-ku, Chiba 263. those in the higher energy region have not been completely established so far. Some computer codes such as High Energy Transport Code (HETC)(3) and Nucleon Meson Transport Code (NMTC)(4) have been utilized for engineering applications of the spallation reaction. These codes contain nuclear calculation routines based on an intranuclear-cascade-evaporation (INCE) model by Bertini(5). Several groups modified the original codes into such versions as HETC-3STEP(6), LAHET(7), HETC/KFA2(8) and NUCLEUS(9)(10). These codes were used both for designing facilities for applications of the spallation neutrons and for estimating nuclear data in the intermediate energy region. Experiments on double-differential cross sections for neutron production have been carried out by the timeof-flight (TOF) method at Indiana University Cyclotron Facility (IUCF)(11) with incident proton energies of 120 and 160MeV, at Paul Scherrer Institute (PSI)(12) with that of 585MeV, and at Los Alamos National Laboratory (LANL)(13)-(18) with those of 113 to 800MeV. Results of intranuclear-cascade-evaporation calculations 529

2 530 K. ISHIBASHI et al. were compared with the experimental data(11)-(18) and some drawbacks of calculation codes were pointed out particularly at incident proton energies of 100 to 200 MeV. At incident energies higher than 800MeV, systematic data covering many targets have not been taken so far. For this reason, we planned to measure the neutronproduction double-differential cross sections from the spallation reaction at the p2 beam line of 12-GeV proton synchrotron (12-GeV PS) at the National Laboratory for High Energy Physics (KEK). For emitted protons with energies above 100MeV, validity of the cascade model has been confirmed(5)(19) to a considerable degree, and this result is expected to hold for produced neutrons. Hence, we were interested in emitted neutrons mainly ranging from 1 to 100MeV. At the p2 beam line, the beam intensity is quite weak, as low as fa level, because protons are generated as secondary particles from an internal target which is placed in the synchrotron accelerator ring. Therefore, a preliminary experiment(20) was performed to find out an experimental method suitable for the measurement in the beam line. In this work, we describe the results of the primary experiment on the neutron-production doubledifferential cross sections. They were obtained on targets of C, Al, Fe, In and Pb at incident proton energies of 0.8, 1.5 and 3.0GeV. Comparisons are made between the experimental data and the calculation results given by the INCE model, and imperfections remaining in the model are discussed. II. EXPERIMENT Since the experimental method was written in our previous report(20), a brief description is made here. The experimental arrangement is shown in Fig. 1 with a simplified illustration. At the p2 beam line of KEK, the beam intensity was very weak in a level of 105 particles/2.5s: The particles came as a low intensity pulsed beam so that they were able to be individually counted. Incident protons were accompanied with pions having the same momentum, due to the nature of the secondary beam. A time-of-flight (TOF) method with a pair of Pilot U scintillators was adopted for the separation of incident protons and pions. A time resolution of 0.25ns was achieved for the incident-particle TOF, and pions were well separated by electronic circuits. NE102A plastic scintillators served to define the incident proton beam area on a target. The coincidence of signals from these scintillators was counted to find the number of incident protons. Weakness of the beam intensity forced us to choose a rather thick target to maintain the measurement efficiency to a reasonable degree. Target characteristics and incident proton energy losses are listed in Table 1, where mean excitation energy stands for the mean excitation and ionization potential in the proton energy loss process in the targets. All targets had a cylindrical shape except for the Pb target which was a 10x10x1.2cm3 plate set at a slanting angle of 15d with respect to a plane perpendicular to the incident beam axis. For the neutron TOF measurement, two different Fig. 1 Illustration of experimental arrangement JOURNAL OF NUCLEAR SCIENCE AND TECHNOLOGY

3 Measurement of Neutron-Production Double-Differential 531 Table 1 Target characteristics and incident-proton energy losses All targets have a cylindrical shape except a Pb plate. sizes of NE213 neutron detectors were placed in directions of 15, 30, 60, 90, 120 and 150d, as shown in Fig. 1. The larger NE213 detectors were 12.7cm in diameter and 12.7cm thick, and were mounted on a plane vertical to the floor. Since the detectors possess a relatively higher detection efficiency, they were used for detecting neutrons with energies higher than 3MeV. The smaller NE213 detectors have a size of 5.08cm in diameter and 5.08cm thick, and they were set on a horizontal plane. The smaller detectors were adopted to measure low energy neutrons in the energy region of 1 to 3MeV, because they exhibited better characteristics of pulseshape discrimination between neutrons and gamma rays in the low energy range. Flight path lengths were 1 to 1.5m for the larger detectors and 0.6 to 0.9m for the smaller ones. In front of the individual neutron detectors, NE102A plastic scintillators of 1cm thick were mounted as veto detectors to eliminate charged-particle events by the anti-coincidence method. A block diagram of measurement circuits is shown in Fig. 2 with simplified drawing. When incident particles came on the target, they brought about the coincidence of signals from all beam detectors at a coincidence module (COIN. 1). Then, a pulse with a typical time duration of 150ns was generated by a gate generator (G.G.1), and was sent to the next coincidence module of COIN. 2. If a signal from neutron detectors reached COIN. 2 through a fan-in module (FAN IN) during the above gate time of 150ns, start pulses were sent to the next gate generators to actuate CAMAC modules such as analog-todigital converters (ADCs) and time-to-digital converters (TDCs). TDCs were used for measuring the TOF that corresponded to the time difference between the stop pulses from the neutron detector and the last incident proton beam detector (POL). The anode pulses from the photomultipliers of the NE213 neutron detectors were branched into four by signal dividers and the electronic charges of the three branches were recorded by charge sensitive ADCs. ADCs 1 and 2 worked for the pulseshape discrimination(20)(21) by prompt and delayed gates with widths of 40 and 350ns, respectively, while ADC3 was used for determining the threshold level with a Fig. 2 Block diagram of measurement circuit with simplified drawing VOL. 34, NO. 6, JUNE 1997

4 532 K. ISHIBASHI et al. total gate width of 250ns. Signals from the veto detectors were sent to ADC4 operated by the same gate width as in ADC3. A m-vax computer was utilized for controlling CAMAC modules. All experimental data were recorded on magnetic tapes and then analyzed by an off-line method. III. DATA ANALYSIS Both target-in and -out measurements were performed in the experiment. Neutron spectra were obtained by subtracting the results of the target-out measurement from those of the target-in, after normalization with the number of incident protons. Examples of TOF spectra of the target-in and -out measurements for 3.0-GeV protons incident on In are shown in Fig. 3, where the spectra include events of both neutrons and gamma rays. Flash gamma rays were generated from target nuclei excited by incident protons. The flash gamma-ray peak was taken as the time reference for the neutron TOF. Target-out results also show flash gamma rays, which are produced by upstream plastic scintillators, and slightly prior to those from the target. The target-in spectra have a broad peak at a time of 30ns. The broad peak indicates neutrons from the evaporation process. The Neutron TOF spectra were separated from gamma rays by pulse-shape discrimination(20). The numbers of neutron-detection events were converted into the double-differential cross sections by using efficiencies of the neutron detectors in the same way as in our previous paper(20). The neutron detection efficiencies were obtained from calculation results of SCINFUL(22) and CECIL(23) codes. The SCINFUL code is applicable to NE213 and NE 110 scintillators for neutrons below 80MeV, whereas CECIL predicts fairly well the neutron detection efficiencies for those up to several hundred MeV. Hence, the results of SCINFUL were utilized for neutron detection efficiencies below 80MeV. The results of CECIL were adjusted to smoothly connect with those of SCINFUL at 80MeV and were employed above 80MeV. The calculation results of the neutron detection efficiency with four bias levels are shown in Fig. 4. The 241Am bias was used only for the smaller NE213 detectors of 5.08x5.08cm2 and the other biases were used for the larger detectors of 12.7x12.7cm2. The energy ranges that are covered by the individual biases are indicated near the upper abscissa. Constant fraction discriminators (CFDs) were utilized for the neutron TOF measurement. The CFD produces a logic pulse at a constant fraction of an original pulse height, and its timing is in principle independent of the pulse height. In actual use, however, timing tends to be faster with increasing pulse height, resulting in a time walk effect in the TOF. This is ascribed to the signal saturation which occurs in photomultipliers for high energy radiation detection. The time walk effects were checked by making two-dimensional plots on ADC3 and TDC for the flash gamma rays. The TOF was found to advance up to 1ns with increasing pulse heights. The TDC values for neutrons were corrected depending on the channels of ADC3. The influence of this correction was significant in the time resolution of the TOF for neutrons above several hundred MeV. The targets used for our measurement were not as thin as those of other measurements(11)-(18). The influence of multiple-scattering neutrons may increase considerably with target thickness. The effect of multiple-scattering was quantitatively checked by the calculation with combination of NMTC/JAERI(24)(10) and MCNP4A(25) codes. The MCNP4A code was adopted to take into consideration the interaction of Fig. 3 TOF spectra of target-in and -out measurements obtained by the 12.7x12.7cm2 NE213 detector at 30dfor 3.0GeV protons on In The spectra include both neutron and gammaray events. The values of TOF are set to zero at the peak of flash gamma rays. The upper scale indicates the energy of neutrons. Fig. 4 Calculated neutron detection efficiencies for NE213 liquid scintillators Results of CECIL are normalized to smoothly connect with those of SCINFUL at 80MeV for the AmBe bias. JOURNAL OF NUCLEAR SCIENCE AND TECHNOLOGY

5 Measurement of Neutron-Production Double-Differential 533 secondary neutrons below 20MeV. The code was used with a continuous energy cross section library, FSXLIB- J3R2(26), which was processed from the nuclear data of JENDL-3.1. The cross section obtained for the ideal thin target was divided by the apparent cross section calculated for the actual thick target. The ratio reaches 0.5 at energies below a few MeV, while it is around unity above 10MeV. The experimental data are multiplied by the ratios calculated at individual energies, that is, the ratios work as the multiplication factors for data correction. Table 2 summarizes typical values of uncertainties of multiplication factors for obtaining the differential cross sections. The uncertainty for the correction of the multiple-scattering was decided to be 20%. The uncertainty of the neutron detection efficiency was deterrnined(20) to be 10% in the neutron energy region below 80MeV, while it was set at 15% at energies above 80MeV. In usual TOF experiments with pulsed beams, the time resolution is mainly determined by the pulse width of incident beams, and it amounts to a few to several ns. In contrast, the incident protons were individually counted in this TOF measurement. The uncertainty of the neutron TOF, therefore, was dominated by the time resolution of neutron detectors. The time resolution of the detectors was obtained from FWHM of the flash gamma-ray peak. The time resolution after excluding the effect of uncertainty of flight path length was evaluated to be 0.5 to 1ns, depending on the amount of the time walk correction. The time resolution was converted into the energy resolution by taking into account the uncertainty of flight path length. The energy resolution of one sigma is shown in Fig. 5. The typical resolution is 5% at a neutron energy of 10MeV and 8% at 100MeV. The energy resolutions in our measurement are sufficiently good in the energy region of 1 to 100MeV with which we were mainly concerned. IV. RESULTS AND DISCUSSION Table 2 Multiplication factors for data correction and their estimated uncertainties Fig. 5 Energy resolution of TOF measurement Double-differential neutron-production cross sections for C, Al, Fe, In and Pb are presented in Figs. 6 to 10, respectively. Dashed lines show the calculation results of the HETC(3) code making use of free nucleon-nucleon elastic scattering cross sections. Solid lines indicate the calculation results of NUCLEUS(10) which is the nuclear reaction computation part of the particle transport code NMTC/JAERI. NUCLEUS takes into account the inmedium effect(27) in target nuclei: the code adopts the modified nucleon-nucleon elastic scattering cross sections of Cugnon(28)-(30) instead of the free nucleon-nucleon elastic scattering cross sections in HETC. In these figures, error bars indicate all experimental uncertainties in one sigma including statistical errors. Dotted lines at the incident energy of 800MeV indicate the experimental data at LANL(17). Our results are mostly in good agreement with the LANL data for Al to Pb. For neutron energies above 200MeV at middle and backward angles, the present cross sections are larger than the LANL data. This is ascribed to the time walk effects that are not corrected for to some extent. There is some deviation for the cross sections for C. However, the present data for C have a tendency to be closer to the calculation results of HETC than the measurement data of LANL. In Figs. 6(a) to 10(a) for the 0.8-GeV proton incidence, the HETC calculations tend to give higher cross sections than the experiments in the neutron evaporation region of 1 to 10MeV, The overestimation is more appreciable for target nuclei lighter than In. However, NUCLEUS generally reproduces those experimental data. The better agreement obtained by NUCLEUS is ascribed to adoption of the in-medium nucleon-nucleon elastic scattering cross sections. Since the in-medium cross sections are considerably smaller than the free ones particularly at energies below 100MeV, nucleons after cascade collision are emitted from nuclei more readily in the NUCLEUS calculation. This reduces the excitation energy of the residuals, resulting in suppression of the neutron emission in the subsequent evaporation process. In the neutron energy range of 10 to 20MeV, the calculations of HETC for Al and Fe overestimate mostly twice the cross sections at the incident energy of 0.8GeV, while the overestimation is removed in those of NUCLEUS. This neutron energy range exists in the overlapping region of the cascade and the evaporation processes. The improvement of NUCLEUS is also ascribed to adoption of in-medium correction for nucleon-nucleon collision. With increasing incident proton energy, the results of VOL. 34, NO. 6, JUNE 1997

6 534 K. ISHIBASHI et al. Fig. 6 Neutron-production double-differential cross section for C results by HETC and NUCLEUS, respectively. Fig. 7 Neutron-production double-differential cross section for Al results by HETC and NUCLEUS, respectively. JOURNAL OF NUCLEAR SCIENCE AND TECHNOLOGY

7 Measurement of Neutron-Production Double-Differential 535 Fig. 8 Neutron production double differential cross section for Fe results by HETC and NUCLEUS, respectively. Fig. 9 Neutron-production double-differential cross section for In results by HETC and NUCLEUS, respectively. VOL. 34, NO. 6, JUNE 1997

8 536 K. ISHIBASHI et al. NUCLEUS become larger than the experimental data at the neutron energies of 10 to 30MeV, and the overestimation is mostly a factor of two at the proton energy of 3GeV for all targets except C. At incident proton energies above 1GeV, the fragmentation reaction(31), which is capable of emitting such nuclei as Na and Mg, is considered to occur frequently after the cascade process. The emission of fragments leads to decrease in the excitation energy of the residuals. Since neither the original NMTC nor HETC includes this reaction, the neglect of the fragmentation may cause overestimation of the neutron emission from the evaporation process. For neutrons with energies above 100MeV, experimental data were obtained mainly at forward and middle angles. The experimental data are reproduced by the calculations in general, particularly at 15d. The neutrons above 100MeV are emitted in the cascade process, and those are not sensitive to the use of the in-medium corrected nucleon-nucleon collision in calculations. Discrepancies between the calculated results and the experimental data are seen at the highest neutron energies in some directions, for example, 60d and 90d. This is because the flight path length was as short as 1m and the time walk correction was not complete in such cases. V. CONCLUSIONS Neutrons were measured for the proton-induced spallation reaction by the TOF method with a typical flight path length of 1m. Neutron-production doubledifferential cross sections were obtained for the targets of C, Al, Fe, In and Pb at proton energies of 0.8, 1.5 and 3.0GeV with energy resolutions better than 8% at neutron energies below 100MeV. Multiple-scattering and time-walk effects were influential in data corrections of the cross section and energy resolution. At the incident proton energy of 0.8GeV, the INCE model calculation with the in-medium nucleon-nucleon elastic scattering cross sections reproduces the experimental data better than that with the original free nucleon-nucleon cross sections. For the incident proton energies above 1GeV, however, comparison between the INCE model calculations and the experimental data suggested that it is required to incorporate the fragmentation reactions into the calculation model. Fig. 10 Neutron production double differential cross section for Pb results by HETC and NUCLEUS, respectively. ACKNOWLEDGMENTS The authors express their gratitude to Prof. Y. Yoshimura, Prof. K. Nakai and the beam channel staff of KEK for their continuous encouragement and generous support of this experiment, and tender their acknowledgments to Prof. H. Hirabayashi and Prof. N. Watanabe of KEK and Prof. M. Arai of Kobe University for their useful discussion. JOURNAL OF NUCLEAR SCIENCE AND TECHNOLOGY

9 Measurement of Neutron-Production Double-Differential 537 -REFERENCES- (1) Rose, R.F.: BNL-NCS-17541, (1991). Shib(2) ata, K.: JAERI-1319, (1990). 3) Chandler, K.C., Armstrong, ( T.W.: CCC-178, (1977). 4) Coleman, W.A., Armstrong, T.W.: ORNL-4606, ( (1970). (5) 6) Bertini, H.W.: Phys. Rev., 188, 1711 (1969). ( Yoshizawa, N., et al.: J. Nucl. Sci. Technol., 32, 601 (1995). (15) 16) Meier, M., et al.: Nucl. Sci. Eng., 110, 289 (1992). ( Amian, W.B., et al.: Nucl. Sci. Eng., 115, 1 (1993). (17) Amian, W.B., et al.: Nucl. Sci. Eng., 112, 78 (1992). (18) Starner, S., et al.: Phys. Rev., C47, 1647 (1993). (19) Harp, G.D.: Phys. Rev., C10, 2387 (1986). (20) Nakamoto, T., et al.: J. Nucl. Sci. Technol., 32, 827 (1995). (21) Nakamoto, T., et al.: Rev. Sci. Instrum., 66, 5327 (1995). ( 22) Dickens, J.K.: ORNL-6452, (1988). (7) Prael, P.E., Lightenstein, H.: Users Guide to the CODE HETC S(23) Cecil, R.A., et al.: Nucl. Instrum. Methods, 161, 439 ystem, Los Alamos National Laboratory (1986). (24) Nakahara, Y., Tsutsui, T.: JAERI-M, , (1982), (1979). (8) Cloth, P., et al.: HERMES, High Energy Radiation M [in Japanese]. ( onte Carlo Elaborate System, KFAIRE-E AN/12/88, 25) Briesmeister, J.F., et al.: LA-12625, (1993). (1988). (26) Kosako, K., et al.: JAERI-Data/Code, (1994). ( (9) Nishida, T., et al.: JAERI-M, , (1986). ( 27) Suetomi, E., et al.: Phys. Lett., B333, 22 (1994). 10) Takada, H.: J. Nucl. Sci. Technol., 33, 275 (1996). (28) Cugnon, J., et al.: Nucl. Phys., A352, 505 (1980). (11) Scobel, W., et al.: Phys. Rev., C41, 2010 (1990). (29) Cugnon, J.: Phys. Rev., C22, 1885 (1980). ( (12) Cierjacks, S., et al.: Phys. Rev., C36, 1976 (1987). ( 30) Niita, K., et al.: Phys. Rev., C52, 2620 (1995). 13) Amian, W.B., et al.: Proc. Nucl. Data for Sci. and (31) Lynch, W.G.: Ann. Rev. Nucl. Part. Sci., 37, 493 Technol., Julich, Germany, 1991, p.696 (1991). (1987). (14) Meier, M., et al.: Nucl. Sci. Eng., 102, 310 (1989). VOL. 34, NO. 6, JUNE 1997