Initialize data. read input file. particle in system? START. find the collided. photon. neutron. nuclide and its in bank? in bank? interaction.

Transcription

1 Application of EDF uclear Data for Testing a Monte-Carlo eutron and Photon Transport Code P. Siangsanan, W.Dharmavanij and S. Chongkum Physics Division, Office of Atomic Eneegy for Peace (OAEP), Ministry of Science Technology and Environment, Thailand A Monte-Carlo photon and neutron transport code was developed at OAEP. The code was written in C and C++ languages in an object-oriented programming style. Constructive solid geometry (CSG), rather than combinatorial, was used such that making its input file more readable and recognizable. As the first stage of code validation, data from some EDF files, in the MCP's specific format, were used and compared with experimental data. The neutron (from a 00 mci Am/Be source) attenuation by water was chosen to compare the results. The agreement of the quantity 1=± among the calculation from SIPHO and MCP, and the experiment which are 10.9 cm, 9.71 cm and cm respectively was satisfactorily well within the experimental uncertainties. These results also agree with the 10.8 cm result of.m., Mirza, et al. 1. Intrduction We have developed our own Monte-Carlo particle transport code. The particles of our interest are neutron and photon, because they are available in our laboratory. We have used the reputed MCP code, including its ACE nuclear data libraries, as our benchmark. Obviously, the MCP is very long and complicated code because it was written in FORTRA and has an old-style programming. It could have taken us much effort to study MCP's source code directly. Instead, we have chosen to study the calculation method from other sources (including MCP manual, of course) and developed our own code using MCP as a guide. We chose C/C++ language because it has a neat structure, especially the OOP styleofc++. ot only the enhanced speed of code development, its source code reuseability is also important. We can add new features to the code with much less effort comparing to FORTRA. Some people may worry about speed penalty to choosing C/C++ over FORTRA, but we do not, because there are many good numerical libraries written in C/C++ that is comparable to that in FORTRA. In our case wehavechosen BLITZ++ (from because the vector operations are extensively used in our code. 2. The Code The name of the code is Simple Implementation of PHOton and eutron transport code (SIPHO). It is not final yet, but it has some major useful features. We have used the concept of Constructive Solid Geometry (CSG)[8] which define the geometry as a solid object. There are four basic shapes in SIPHO: cuboid, cylinder, sphere and cone. Objects of other shape could be obtained through Boolean operation between them. Some features of SIPHO are: ffl User interface via text input file (an example input file in this benchmark is shown below) ffl Repeated structure ffl Fixed particle sources ffl eutron and photon transport ffl Thermal neutron scattering treatment ffl Photon energy deposition in a cell ffl Track length estimate of particle flux in a cell The flowchart of the code is shown in fig.1. 1

2 Table 1: An example of input file objectf #declare srcwall = cylinderf<0,0,-.9>, <0,0,.9>,.87g mat<0> g #declare srcvial1 = cylinderf <0,0,-1.1>, <0,0,1.97>,1.1 g #declare srcvial2 = cylinderf <0,0,-.9>, <0,0,1.77>,.9 g objectf srcvial1 ~ srcvial2 mat<7> g #declare srctube = objectf #declare srcvoid1 = cylinderf <0,0,-15>, <0,0,2>, 2.g ~ #declare srcvoid2 = cylinderf <0,0,-14.8>, <0,0,2>, 2.1 g mat<4> g objectf srcvoid2 ~ srcvial1 mat<6> g objectf srcvial2 ~ srcwall mat<6> g #declare detvshroud = cylinderf <16.2,0,-1.5>, <16.2,0,1.5>, 1.1 g #declare uppercd1 = objectf #declare ucd1 = cylinderf <16.2,0,-1.5>, <16.2,0,-1>, 1.49 g ~ cylinderf <16.2,0,-1.5>, <16.2,0,-1>, 1.1 g mat<2> g #declare lowercd1 = objectf #declare lcd1 = cylinderf <16.2,0,1>, <16.2,0,1.5>, 1.49g ~ cylinderf <16.2,0,1>, <16.2,0,1.5>, 1.1g mat<2> g #declare knifetool1 = cylinderf <16.25,2,-14>, <16.25,2,14>, 2.1g #declare knifetool2 = cylinderf <16.25,-2,-14>, <16.25,-2,14>, 2.1g #declare knifetool = cylinderf <14.25,0,-14>, <14.25,0,14>, 2g #declare patchcd1 = objectf cylinderf <16.2,0,-1.5>, <16.2,0,-1>, 1.58g ~ ucd1 ~ knifetool1 ~ knifetool2 ~ knifetool mat<2> g #declare patchcd2 = objectf cylinderf <16.2,0,1>, <16.2,0,1.5>, 1.58 g ~ lcd1 ~ knifetool1 ~ knifetool2 ~ knifetool mat<2> g #declare dettube = objectf #declare detvoid1 = cylinderf <16.2,0,-15>, <16.2,0,2>, 1.9g ~ #declare detvoid2 = cylinderf <16.2,0,-14.8>, <16.2,0,2>, 1.7g mat<4> g objectf detvoid2 ~ uppercd1 ~ lowercd1 ~ patchcd1 ~ patchcd2 ~ detvshroud mat<6> g #declare detwall = objectf detvshroud ~ #declare detel = cylinderf <16.2,0, >, <16.2,0,10.795>, g mat<8> g #declare detector = objectf detel mat<5> g #declare splittank1 = cuboidf <-55,-55,-55>, <5.25,55,55> g #declare splittank2 = cuboidf < 5.25,-55,-55>, <8.25,55,55> g #declare splittank = cuboidf < 8.25,-55,-55>, <11.25,55,55> g #declare splittank4 = cuboidf < 11.25,-55,-55>, <55,55,55> g #declare tank = cylinderf <0,0,-16>, <0, 0,16>, 5g #declare watertank1 = objectf (tank & splittank1) ~ srcvoid1 mat<1> g #declare watertank2 = objectf tank & splittank2 mat<1> g #declare watertank = objectf tank & splittank mat<1> g #declare watertank4 = objectf (tank & splittank4) ~ detvoid1 mat<1> g objectf #declare otank = cylinderf tank mat<8> g objectf #declare env = spheref <0,0,0>,50 g ~ otank ~ srcvoid1 ~ detvoid1 mat<6> g objectf spheref <0,0,0>,55 g ~ env mat<0> g //+++ Last(data) Section mode = neutron source=fvolume<1>, vect=<1,0,0>, dir=<0>, erg=<d 1> g //Am/Be neutron energy distribution. cdf 1fi=[ ], p=[

3 -cont ]g mat 1[ lwtr.01t] //H2O mat 2[ ] //Cd mat [1.26e ] //B-10 mat 4[ ] // Fe mat 5[1.2875e ] //BF- mat 6[ ] // Air mat 7[ ] // Polyethylene mat 8[ ] // ss04 tallyf neutron flux within cell<detector> g tallyf neutron flux within cell <detector> multiplier<1,,107> g run particles imp[ ] START read input file Initialize data particle in system? find the collided photon neutron nuclide and its in bank? in bank? interaction photon running 1 calculate neutron production? mode? parameters then store in the bank running 1 neutron more than mode? production? 1 neutron? calculate photon parameters then scoring store in the bank Figure 1: The flowchart of SIPHO. Experiment The neutron attenuation in water was chosen to be the first validation scheme of the code. The experimental setup composed of a 00 mci Am/Be neutron source installed in the middle of a 70-cm-diameter water tank. A cadmium-covered BF detector was installed along side

4 of the source to measure the neutron activity. The distance of the detector to the source is allowed to vary from 2 cm to 26 cm with 2-cm-step increment. The thickness of cadmium is 1.8 mm and a 2-cm-wide window was provided at the middle of the detector. The window of the detector was aligned to the center of the source. The setup is shown in fig.2 and the detailed description of the detector is shown in fig. 70 cm 4.6 cm diam. 2 mm thick iron tube r Preamplifier 70 cm.8 cm diam. 2 mm thick iron tube HV supply 00 mci Am/Be neutron source MCA Cannberra S 10 distilled water Cd covered BF detector Stainless steel tank Figure 2: The experimental setup BF detector Detector Properties: 1.8 mm thick cadmium 2 cm window Overall length cm Sensitive length cm Outer shell...ss04, cm Fill pressure...40 cm Hg Fill gas...bf enriched to 96% B 10 Figure : The detector description 4. Calculation The model of the calculation at r = 16.2 cm is shown as the input file in table 1. In this calculation we have treat the implicit neutron capture, thermal neutron scattering and geometry splitting. At other r's the models are similar. 5. Results The results from the experiment and the calculation are shown in table 2. The plot of C B r 2 against r, where C B is the activation of boron within the detector and can be 4

5 calculated from C B = R 1 0 ± (n;ff) (E)ffi(E)dE, is also shown in fig.4. The source emission rate used was 96% of 6: n/s. r counting rate (n/s) C B = R 1 ± 0 (n;ff) (E)ffi(E)dE (cm) SIPHO MCP Experiment C B (rel.error) fom C B (rel.error) fom mean(rel.error) (0.0015) (0.024) (0.00) (0.01) (0.015) (0.006) (0.01) (0.018) (0.004) (0.011) (0.021) (0.0051) (0.01) (0.017) (0.0061) (0.0078) (0.015) (0.0075) (0.014) (0.016) (0.009) (0.016) (0.012) (0.01) (0.015) (0.01) (0.01) (0.024) (0.016) (0.016) (0.02) (0.02) (0.018) (0.02) (0.02).4 6.5(0.02) (0.026) (0.029) (0.027) Table 2: The counting rate from measurements and calculations Experiment, 1/Σ = cm SIPHO, 1/Σ = 10.9 cm MCP, 1/Σ = 9.71 cm C B r 2 (arbitrary unit) r (cm) Figure 4: Plot of C B r 2 against r to determine and compare 1/± 6. Conclusion It was seen from the plot in fig.4 that all curve have the same shape, though they are not exactly aligned. The determination of 1=± using the relation C B ο e ±r =r 2 or 5

6 r 2 C B ο e ±r at r > 1 cm, was found to be 10.9 cm for SIPHO, 9.71 cm for MCP and cm for the experiment. These values agree with 10.8 cm from the work of Mirza,.M., et al.[]. It should be noted that the Monte-Carlo efficiency of SIPHO is rather good, as shown by the figure of merit (fom) in the third column of table 2. However, there are somewhat difference between the result of SIPHO and MCP, which, unfortunately, we have not analysed in detail yet. Acknowledgements We would like to thank the free software community who contributed to every software we have used, especially BLITZ++, in our work. References [1] Lux, I. and L. Koblinger, Monte Carlo Particle Transport Methods: eutron and Photon calculations", CRC Press, Inc., Boca Raton, Florida. 517 p. (1990) [2] Cashwell, E. D. and C. J. Everett, Monte Carlo Method for Random Walk Problems", Pergamon Press, London. 15 p. (1959) [] Mirza,. M., S. M. Mirza and M. Iqbal, Determination of Mean Squared Slowing-down Distance for Am/Be eutrons in Water using BF -Detector", Radiat. Phys. Chem. vol. 48 no. 4 pp (1995) [4] Kludge, H. and K. Weise, The eutron Energy Spectrum of a 241 Am-Be(ff,n) Source and Resulting Mean Fluence to Dose EquivalentConversion Factors", Radiation Protection Dosimetry, vol. 2 o. 2 pp (1982) [5] Briesmeister, J. F., MCP A General Monte Carlo -Particle Transport Code Version 4B", Radiation Information Computational Center (RSICC), Oak Ridge, U.S.A., 709 p. (1997) [6] Eckel, B., Thinking in C++", Prentice Hall, ew Jersey, 81p. (1995) [7] Goldstein, H., Classical Mechanics", Addison-Wesley, U.S.A., pp (1980 2nd Ed.) [8] Foley, J. D., et al., Computer Graphics: Priciples and Practice", Addison-Wesley, U.S.A., pp (1992) 6