Abstract
The purpose of neutronic calculations is to determine many principal integral parameters such as effective multiplication factor (keff), reaction rate, spectrum indices, etc. These parameters, are based on several cross sections as well as their uncertainties. However, the uncertainty propagations effect will give, in turn, inaccurate values of these integral parameters. Therefore, the margin of reactor safety can be decreases. In order to minimize these risks, the adjustment of basic nuclear data (Cross section) is required.
Cross sections adjustment techniques consist essentially of using available information from integral experiments to improve the basic nuclear data. In this work, the multi-group cross section adjustment is processed for data which are primordial in neutronic nuclear calculations and their covariance matrices are available in the ENDF files. The adjustment process is based on the uncertainty of keff using the maximum likelihood method (i.e. general least squares method GLLSM). So, several critical benchmarks, their sensitivities matrices and the cross sections covariance matrix are required when using this method. Benchmarks have been taken from the International Handbook of Evaluated Criticality Safety Benchmark Experiments (IHECSBE), their sensitivity matrices and the covariance matrices of the desired cross sections have been calculated by MCNP6 code and NJOY software respectively. This study investigates the effects of the correlation between reactions on the prior and posterior nuclear data uncertainty, adjusted cross-sections and their standard deviations and the adjusted integral parameters (keff). A significant reduction of the a-priori uncertainties and a good convergence of the C/E ratio are observed after adjustment by using the a-posteriori covariance and cross sections data.
•The group cross section uncertainty estimation to the keff in the ENDF/B-VII.1 evaluation using MCNP6.1 and NJOY99.•The cross section adjustment process by using linear least squares method in two cases (correlation and non-correlation).•The covariance matrix adjustment by using linear least squares method.•The comparison between prior and posterior keff in the two cases.