Finite-geometry and polarized multiple-scattering corrections of experimental fast-neutron polarization data by means of Monte Carlo methods

Abstract

After an introductory discussion of various methods for correction of experimental left-right ratios for polarized multiple-scattering and finite-geometry effects, necessary and sufficient formulas for consistent tracking of polarization effects in successive scattering orders are derived. The simplifying assumptions are then made that the scattering is purely elastic and nuclear, and that in the description of the kinematics of the arbitrary Scattering μ only one triple-parameter - the so-called spin-rotation parameter β(μ) - is required. based upon these formulas a general discussion of the importance of the correct inclusion of polarization effects in any scattering order is presented. Special attention is then paid to the question of depolarization of an already polarized beam. Subsequently, the afore-mentioned formulas are incorporated in the comprehensive Monte Carlo program MULTPOL, which has been designed so as to correctly account for finite-geometry effects, in the sense that both the scattering sample and the detectors (both having cylindrical shapes) are objects of finite dimensions located at finite distances from each other and from the source of polarized fast-neutrons. A special feature of MULTPOL is the application of the method of correlated sampling for reduction of the standard deviations of the results of the simulated experiment. Typical data of performance of MULTPOL have been obtained by the application of this program to the correction of experimental polarization data observed in n+ 12C elastic scattering between 1 and 2 MeV. Finally, in the concluding remarks the possible modification of MULTPOL to other experimental geometries is briefly discussed.