RECENT ADVANCES IN PERTURBATIVE METHODS APPLIED TO NUCLEAR-ENGINEERING PROBLEMS

This paper describes the main perturbative methods used in sensitivity analysis. It also presents several applications to calculate sensitiv ity coefficients of interest to nuclear engineering problems. These to pics will cover only the papers presented in the 10th Brazilian Meetin g on Reactor Physics and Thermal Hydraulics - X ENFIR First the matrix formalism of the perturbation theory is applied in a simplified model to study the hot channel of PWR reactors. The mass, linear momentum a nd energy conservation equations and appropriate heat transfer and flu id mechanic correlations describe the discretized system. After calcul ating system's thermal hydraulic properties, the matrix formalism is a pplied and the sensitivity coefficients are determined for each case o f interest. Comparisons between perturbative and direct calculations s how good agreement, which demonstrates that the matrix formalism is an important tool for discretized system analysis. A second application of the GPT methodology to a reliability engineering problem of great p ractical interest follows: that of the analysis of the influence of th e demand rate on the reliability of a process plant equipped with a si ngle protective channel. The application of the GPT approach needs the solution of the system for a few points (reference solutions). The re sults agree very well with those published in the literature. In a thi rd application, the so called differential method was applied to the s ensitivity analysis of the waterhammer effect in hydraulic networks. A s an example, a constant-level tank connected through a pipe to a valv e discharging to atmosphere was considered. The sensitivity coefficien ts for given responses obtained by using both the differential method and the response surface generated by the computer code WHAT were calc ulated. The results obtained show excellent agreement. As a fourth app lication, the GPT formalism is used to analyze the sensitivity of some responses of interest to boron neutron capture therapy. Attention is focused on diffusive source-driven systems. Finally a methodology is d escribed relevant to the use of data from experimental facilities with respect to a reference design The method makes use of GPT methods for the calculation of the sensitivity coefficients relevant to the respo nses of interest with respect to the system parameters. Adjustment and data transposition methods extensively used in the reactor physics ne utronic domain are considered as applicable to other fields of interes t, in particular to the thermal-hydraulic field. (C) 1997 Elsevier Sci ence Ltd.