Status report on SHARP coupling framework.

Caceres, A. ; Tautges, T. J. ; Lottes, J. ; Fischer, P. ; Rabiti, C. ; Smith, M. A. ; Siegel, A. ; Yang, W. S. ; Palmiotti, G.

关键词: ACCURACY;    ALGORITHMS;    ANL;    DESIGN;    FAST REACTORS;    PERFORMANCE;    PHYSICS;    REACTORS;    SODIUM;    THERMAL HYDRAULICS;   

DOI  :  10.2172/934903
RP-ID  :  ANL-AFCI-215
PID  :  OSTI ID: 934903
Others  :  TRN: US0804104

美国|英语

来源: SciTech Connect

【 摘 要 】

This report presents the software engineering effort under way at ANL towards a comprehensive integrated computational framework (SHARP) for high fidelity simulations of sodium cooled fast reactors. The primary objective of this framework is to provide accurate and flexible analysis tools to nuclear reactor designers by simulating multiphysics phenomena happening in complex reactor geometries. Ideally, the coupling among different physics modules (such as neutronics, thermal-hydraulics, and structural mechanics) needs to be tight to preserve the accuracy achieved in each module. However, fast reactor cores in steady state mode represent a special case where weak coupling between neutronics and thermal-hydraulics is usually adequate. Our framework design allows for both options. Another requirement for SHARP framework has been to implement various coupling algorithms that are parallel and scalable to large scale since nuclear reactor core simulations are among the most memory and computationally intensive, requiring the use of leadership-class petascale platforms. This report details our progress toward achieving these goals. Specifically, we demonstrate coupling independently developed parallel codes in a manner that does not compromise performance or portability, while minimizing the impact on individual developers. This year, our focus has been on developing a lightweight and loosely coupled framework targeted at UNIC (our neutronics code) and Nek (our thermal hydraulics code). However, the framework design is not limited to just using these two codes.