Academic Journals Database
Disseminating quality controlled scientific knowledge

Uncertainty evaluation of nuclear reaction model parameters using integral and microscopic measurements. Covariances evaluation with CONRAD code

Author(s): De Saint Jean C. | Habert B. | Archier P. | Noguere G. | Bernard D. | Tommasi J. | Blaise P.

Journal: EPJ Web of Conferences
ISSN 2100-014X

Volume: 8;
Start page: 04002;
Date: 2010;
Original page

In the [eV;MeV] energy range, modelling of the neutron induced reactions are based on nuclear reaction models having parameters. Estimation of co-variances on cross sections or on nuclear reaction model parameters is a recurrent puzzle in nuclear data evaluation. Major breakthroughs were asked by nuclear reactor physicists to assess proper uncertainties to be used in applications. In this paper, mathematical methods developped in the CONRAD code[2] will be presented to explain the treatment of all type of uncertainties, including experimental ones (statistical and systematic) and propagate them to nuclear reaction model parameters or cross sections. Marginalization procedure will thus be exposed using analytical or Monte-Carlo solutions. Furthermore, one major drawback found by reactor physicist is the fact that integral or analytical experiments (reactor mock-up or simple integral experiment, e.g. ICSBEP, …) were not taken into account sufficiently soon in the evaluation process to remove discrepancies. In this paper, we will describe a mathematical framework to take into account properly this kind of information.

Tango Rapperswil
Tango Rapperswil

RPA Switzerland

RPA Switzerland

Robotic process automation