The Temperature Zoning Method for the On-the-Fly Thermal Scattering Sampling Procedure

Le résumé fourni par la source

Recently, we have developed a methodology to sample a neutron’s secondary energy and flight angle after a scattering collision adaptively in temperature in the thermal energy range [1]. This on-the-fly (OTF) method avoids the traditional sampling procedure that requires pre-storing many data at discrete temperatures. For simulations involving a large temperature variation and/or temperature feedback, the amount of pre-processing work required can be large. The new sampling procedure takes place without recourse to the single-temperature data files. Instead, the temperature dependence of the energy (β) and momentum (α) transfers are used to directly obtain the scattering parameters for thermal neutron scattering simulations. In order to implement this method, the probability density functions (PDFs) of the energy and momentum transfers after a collision are studied and fits are generated at cumulative distribution function (CDF) lines that best describe the temperature change using a least squares approach. The coefficients of these fits are used to perform the sampling. We have previously shown that polynomial fits using a 1/T basis function give the best results while only requiring a few megabytes of data storage [2]. While integral quantities like keff agree well (within 1-2σ) between the standard and OTF sampling methods, problems involving only a few scatters before being tallied perform poorly even if the order of the fitting is increased. This is seen for differential quantities like the secondary energy distribution. Recent work has been performed on using Monte Carlo to calculate multi-group scattering kernels for deterministic codes [3]. Incorrect sampling of the secondary thermal neutron energy when using the OTF method could result in a misrepresentation of these scattering kernels. To improve the simulation accuracy for differential quantities, we explored a different strategy for the fitting. Instead of going for higher order in a single temperature zone, we use a lower order fitting in multiple temperature zones. The work described in this paper compares the results of generating fit coefficients by 1) using a single temperature range and 2) using various temperature zones. The former is the method that we have used in our previous studies while the latter is the new work. This new work is not applied to the secondary angle distributions since no issues have been observed with the OTF sampling of the scattering angle. The examples and results described throughout this paper use bound carbon in graphite as the material of interest.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.