The problem to account for processes on different time scales in the investigation of neutron embrittlement


The problem to account for processes on different time scales in the investigation of neutron embrittlement

Gokhman, A.; Bergner, F.; Ulbricht, A.

Neutrons with energy of about 1 МeV or higher, to which reactor pressure vessel steels are exposed under service conditions of nuclear power plants, can stimulate the phenomenon of so-called neutron embrittlement. The reason is the irradiation-induced formation and evolution of precipitates and solute/defect clusters. The set of experimental methods to investigate the nanostructure, including X- ray and neutron small angle scattering, atom probe field ion microscopy method, TEM as well as direct determination of the mechanical properties of the reactor vessel is usually carried out after irradiation intervals of the order of one year.
Therefore a project called virtual test reactor has been started recently within the framework of the European program “PERFECT”. Experimental data obtained for neutron irradiated pressure vessel steels are used as the reference for working out and study theoretical models of the kinetics of neutron embrittlement.
In the cascade stage, following the interaction of a neutron with a so-called primary knock-on atom, the recoil energy transferred to the crystal lattice causes a local increase of the temperature to beyond the melting temperature and the formation of defects. A typical time scale for these processes is picoseconds. Such processes are studied by the molecular dynamics and Monte-Carlo methods.
Unlike the case of electron irradiation, where only Frenkel pairs are formed, point defect clusters and Cu-precipitates are produced in the case of neutron irradiation. The evolution of these clusters can be studied in the rate theory approach, in particular by cluster dynamics method. This method allows for the determination of the time dependence of the number density of clusters (precipitates) via integration of the system of ordinary differential equations with the initial data as the final data of the cluster number density after cascade stage. The characteristic relaxation time depends on the kind of the irradiated feature. For example for point defect clusters this time is about 1 day but for Cu-precipitates the relaxation time is about one month. Hence the problem to integrate the master equation of cluster dynamics is the problem of integration of so-called stiff ordinary differential equations.
The objective of our present investigation is the comparative analysis of software tools applied to the cluster dynamics modeling. The most effective one, namely the code D02EJF from the licensed Fortran NAG Library has been used to investigate the kinetics of the following three kinds of neutron irradiated features in reactor pressure vessel steels:

  • The system of free and clustering vacancies (V) and interstitials (I), denoted (V+I)-system,
  • The system of the Cu-precipitates favoured because of irradiation-enhanced Cu diffusivity, denoted (V+I)=>Cu system,
  • The (V+I)óCu system, when the effect of Cu-precipitates on the sink strength in the first and second model is additionally considered.
The comparison of the experimental and cluster dynamics data has been carried out on the basis of low-Cu and Cu-enriched mock-up alloys after neutron irradiation.
  • Lecture (Conference)
    4-th Russian Symposium "Physics Problems of Ultrashort Phenomena in Strongly Non-Equilibrium Media", 22.07.-01.08.2006, New Athos, Russia

Permalink: https://www.hzdr.de/publications/Publ-8608
Publ.-Id: 8608