The Method of Characteristics (MOC) is widely applied to high-fidelity numerical simulations due to its robust geometric processing capabilities, as well as its ability to balance computational costs and accuracy during calculations. In addition to MOC, common neutron transport calculation methods also include the Collision Probability method (CP) and the Interface Current method (IC). In MOC calculation, different parameter selections will lead to different values of calculation cost and accuracy.

This paper aims to evaluate the ability of MOC, CP and IC methods for pin-by-pin calculation, and conduct sensitivity analysis to find the best parameter setting for MOC method.

The three aforementioned calculation methods were compared from the perspective of theory and numerical calculation. Subsequently, numerical calculation and preliminary analysis of the sensitivity of MOC parameters were conducted based on the 2D C5G7-MOX reference problem.

Numerical calculation results show that the computation time and memory cost incur by the MOC are 23.9 min and 37.5 MB, respectively, and the relative error between the MOC results and reference solutions is only 6.04×10^-4. The computing times of the CP and IC methods are 56.7 times and 15.6 times that of the MOC, and the memory costs are 407.7 times and 32.8 times that of the MOC, respectively. As a result of the sensitivity analysis of MOC parameters, the following set of parameters is suggested: a grid division of 6×6, a pole angle of GAUS, a pole number of 2, and an azimuth angle of 30°. The calculation time and the memory cost of this set of parameters are 45.4 min and 264.7 MB, respectively, with the relative error of 5.9×10^-5 and the normalized RMS error of 0.002 55.

The results of this study indicate that the MOC is superior to the CP and IC methods in accuracy, efficiency, and memory cost, and the grid division of MOC has the greatest influence on the calculation memory cost and calculation time whereas the choice of polar angle has the greatest influence on the calculation accuracy. With its powerful geometric processing ability and consideration of the calculation cost and accuracy, the MOC is more widely used in high-fidelity numerical simulation for neutron transport calculation.