Optimization of GUARDYAN for NPP use

Nyomtatóbarát változatNyomtatóbarát változat
Doctoral school: 
Fizikai Tudományok Doktori Iskola
Year/Semester: 
2022/2023/1
Supervisor
Name: 
David Legrady
Email: 
legrady@reak.bme.hu
Institute: 
Institute of Nuclear Techniques
Job title: 
Associate Professor
Academic degree: 
PhD
Description: 

GUARDYAN (GPU Assisted Reactor Dynamic Analysis) is a continuous energy Monte Carlo (MC) neutron transport code developed at BME NTI. It targets to solve time-dependent problems related to fission reactors with the main focus on simulating and analyzing short transients. The key idea of GUARDYAN is a massively parallel execution structure making use of advanced programming possibilities available on CUDA (Compute Unified Device Architecture) enabled GPUs (Graphics Processing Units). It has recently been shown that GUARDYAN can be upscaled to realistic Nuclear Power Plant (NPP) geometries, including burnup, temperature, and density distribution to a level when real ex-core detector measurement can be reproduced. Though the demonstration of this capability was successful, running times are still discouraging, especially when because higher grade GPU's underperformed their expectations by order of magnitude to reasons unknown. Further, calculation speed-up techniques that worked very efficiently in small reactor geometries failed to improve calculation times. Coupling to thermal-hydraulic solvers requires optimization of its own, a topic so far unvisited by the GUARDYAN developer group. The Ph.D. topic involves the further development of GUARDYAN to handle realistic NPP geometries efficiently. Existing efficiency-improving techniques should be upscaled to NPP level; New algorithms should be developed to harness higher-level GPU features. Detailed optimization should target the coupling to thermal-hydraulic solvers to achieve ultimate fidelity DMC calculations on an NPP scale within practical calculation time frames.

Requirements: 
profound understanding of Monte Carlo methods, experience with DMC, CUDA programming skills, aptitude for algorithm design,, good understating of reactor physics
Project type: 
PhD project for standard admission
Status: 
Finalized/Végleges