Implementation and Comparison of Contact Models within PIRAT for Nuclear Reactivity Control Systems
Résumé
This work shows the results for a stepped-sine excitation of a Reactivity Control System for a nuclear reactor. These simulations use the PIRAT toolbox, which is currently in development at CEA to study insertion reliability, to predict the maximum displacement of the response location compared to the static estimate used in previous studies. Within the PIRAT toolbox, the dynamic solver (DEBSE) has two contact models implemented to study the interaction between the subsystems of the RCS. One of these is a user-inputted, mass-less, two-stage spring to mimic the fluid compression in the first stage and material contact in the second, while the other model only uses material contact via the Lankarni and Nikravesh model with limited user-inputted parameters. The system is excited based on a stepped-sine excitation with a maximum amplitude to mimic the previous work using a static approach. This sweep shows that each contact model (as well as the same model with varied parameters) can produce significantly different responses. In addition to the varied responses, the different models had difficulty near resonance effects that manifest as numerical ill conditioning. This work shows the need for experimental calibration of these models due to the variability as well as some of the possible issues when performing these calculations. While this work does not present any novel contact modeling, it does show a new implementation for a nuclear energy system and some of the unique issues that arise from these unique systems.
Origine : Fichiers produits par l'(les) auteur(s)
Loading...