Important Cautions in Shielding Computation by using the FLUKA Code

It is important to consider how efficiently the facility is designed against the shielding and activation problems in order to minimize the radiation exposure. In order to evaluate shielding capability of facility, although the simple calculation with approximate methods were popular until recently, owe to the development of information technology and the advances of computational mathematics, the Monte Carlo codes such as the MCNP, FLUKA, GEANT, and PHITS which can provide accurate answer are popularly used now. Advantage of Monte Carlo code is to perform the correct calculation but takes a long time for computing. More importantly, the exact and precise input data for Monte Carlo codes is essential in order to obtain accurate results. Thus, in this paper, important cautions are presented for shielding computation with the FLUKA code because the ignorance of such important cautions makes big troubles. The absorbed doses and errors show similar tendency in the comparison of groups, 1 and 2. Specifically, the results confirmed the more similar tendency in the high convergence areas. In group 3, although the comparison with groups, 1 and 2, shows the similar absorbed dose in the detectors with high convergences, the results themselves are unreliable because the errors are 99.9%. Thus, we need more careful attention to the average value and error in using the FLUKA code. Simply, it is better for us to have other benchmark tools such as MCNPX. However, it is recommended that the best computing method with the FLUKA code is the same as the computing of group 2, the usual multiprocessing with semi-automatic data handling. As shown in group 3, higher number of the cycle is a better method than the higher history to get more reliable result or to reduce errors. However, these values should be carefully evaluated