Zirconium evaluations for ENDF/B-VII.2 for the fast region

We have performed a new combined set of evaluations for ⁹⁰⁻⁹⁶Zr, including new resolved resonance parameterizations from Said Mughabghab for ^{90,91,92,94,96}Zr and fast region calculations made with EMPIRE-3.1. Because ⁹⁰Zr is a magic nucleus, stable Zr isotopes are nearly spherical. A new soft-rotor optical model potential is used allowing calculations of the inelastic scattering on low-lying coupled levels of vibrational nature. A soft rotor model describes dynamical deformations of the nucleus around the spherical shape and is implemented in EMPIRE/OPTMAN code. The same potential is used with rigid rotor couplings for odd-A nuclei. This then led to improved elastic angular distributions, helping to resolve improper leakage in the older ENDF/B-VII.1β evaluation in KAPL proprietary, ZPR and TRIGA benchmarks. Another consequence of ⁹⁰Zr being a magic nucleus is that the level densities in both ⁹⁰Zr and ⁹¹Zr are unusually low causing the (n,el) and (n,tot) cross sections to exhibit large fluctuations above the resolved resonance region. To accommodate these fluctuations, we performed a simultaneous constrained generalized least-square fit to (n,tot) for all isotopic and elemental Zr data in EXFOR, using EMPIRE's TOTRED scaling factor. TOTRED rescales total cross sections so that the optical model calculations are unaltered by the rescaling and the correct competition between channels is maintained. In this fit, all (n,tot) data in EXFOR was used for Ein > 100 keV, provided the target isotopic makeup could be correctly understood, including spectrum averaged data and data with broad energy resolution. As a result of our fitting procedure, we will have full cross material and cross reaction covariance for all Zr isotopes and reactions.