Revisiting mini-max polynomial approximation method for nuclear fuel depletion calculation

Nuclear fuel depletion calculations with detailed nuclide transmutation chains require specific numerical methods and the Chebyshev rational approximation method (CRAM) has been widely used. The mini-max polynomial approximation (MMPA) method is also for fuel depletion calculations and has several advantages over CRAM. The original MMPA coefficients are determined to minimize polynomial approximation errors over an entire range of a variable. In the present paper, relation between the polynomial approximation errors and reproduction errors of nuclide number densities (ND) is carefully investigated, and it is found that the MMPA coefficients which minimize approximation errors in a specific range of the variable can reduce the reproduction errors of ND. Reference NDs are reproduced within 1% differences with the new MMPA coefficients with the 8th order for fuel depletion calculations of PWR-simulated pincells.