No abstract

No abstract

No abstract

This paper presents a global approach to the validation of the parameters that enter into the neutronics simulation tools for advanced fast reactors with the objective to reduce the uncertainties associated to crucial design parameters. This global approach makes use of sensitivity/uncertainty methods; statistical data adjustments; integral experiment selection, analysis and "representativity" quantification with respect to a reference system; scientifically based cross section covariance data and appropriate methods for their use in multigroup calculations. This global approach has been applied to the uncertainty reduction on the criticality of the Advanced Burner Reactor, (both metal and oxide core versions) presently investigated in the frame of the GNEP initiative. The results obtained are very encouraging and allow to indicate some possible improvements of the ENDF/B-VII data file.

iii SUMMARYIn this annual report we illustrate the methodology of the consistent data assimilation that allows to use the information coming from integral experiments for improving the basic nuclear parameters used in cross section evaluation.A series of integral experiments were analyzed using the EMPIRE evaluated files for 235 U, 238 U, and 239 Pu. Inmost cases the results have shown quite large worse results with respect to the corresponding existing evaluations available for ENDF/B-VII.The observed discrepancies between calculated and experimental results were used in conjunction with the computed sensitivity coefficients and covariancematrix for nuclear parameters in a consistent data assimilation. Only the GODIVA and JEZEBEL experimental results were used, in order to exploit informations relative to the isotope of interest that are, in this particular case:235 U and 239 Pu.The results obtained by the consistent data assimilation indicate that with reasonable modifications (mostly within the initial standard deviation) it is possible to eliminate the original large discrepancies on the K eff of the two critical configurations. However, some residual discrepancy remains for a few fssion spectral indices that are, most likely, to be attributed to the detector cross sections.

In this paper, a new neutron kinetics solver for cylindrical R-Z geometry, CYNOD, is presented for the simulation of coupled transient problems for pebble bed reactors. The code utilizes the Direct Coarse Mesh Finite Difference method, in which a set of one-dimensional equations in each transverse direction is solved by means of the analytic Green's function method. A method that deals with control rod cusping problems is also presented. A heterogeneous fuel kernel model is implemented in order to accurately take into account Doppler feedback effects. Numerical results that demonstrate the accuracy of the code are also presented.

No abstract

A new approach is proposed, the consistent data assimilation, that allows to link the integral data experiment results to basic nuclear parameters employed by evaluators to generate ENDF/B point energy files in order to improve them. A practical example is provided where the sodium neutron propagation experiments, EURACOS and JANUS-8, are used to improve via modifications of 23 Na nuclear parameters (like scattering radius, resonance parameters, Optical model parameters, Statistical Hauser-Feshbach model parameters, and Preequilibrium Exciton model parameters) the agreement of calculation versus experiments for a series of measured reaction rate detectors slopes. Future work involves comparison of results against a more traditional multigroup adjustments, and extension to other isotope of interest in the reactor community as 56 Fe, actinides, and fission products.