Author: Publisher:
ISBN:
Category :
Languages : en
Pages :
Book Description
Successful nuclear design requires adequate prediction of integral design parameters, and this in turn requires an adequate differential nuclear data base. Data bases that apparently permit reduced biases and design margins have been developed by a) least squares adjustment of differential data or b) trial-and-error selection from alternative evaluated data sets. Criticisms and defenses of such procedures are discussed. Useful data adjustment is related to consistency of the combined differential-integral data set and consistency tests related to least squares adjustment procedures are described. An approach to data adjustment is suggested that is contingent on consistency analysis. A FORTRAN code ALVIN has been developed to carry out the indicated data consistency and adjustment calculations, and to compute required sensitivities of integral parameters to nuclear data changes. The sensitivity modules of ALVIN are validated by computing with two distinct methods the cross-section sensitivity profile for neutron penetration through a thick iron shield. The data consistency and adjustment modules DAFT2 (for arbitrary variance-covariance data) and DAFT3 (for differential data base of arbitrary size uncorrelated with integral data) are validated by comparing their results for a set of data for three ZPR criticals. 2 figures, 5 tables, 51 references (auth).