1 |
Analysis and Improvement of the bRAPID Algorithm and its ImplementationBartel, Jacob Benjamin 18 July 2019 (has links)
This thesis presents a detailed analysis of the bRAPID (burnup for RAPID – Real Time Analysis for Particle transport and In-situ Detection) code system, and the implementation and validation of two new algorithms for improved burnup simulation. bRAPID is a fuel burnup algorithm capable of performing full core 3D assembly-wise burnup calculations in real time, through its use of the RAPID Fission Matrix methodology. A study into the effect of time step resolution on isotopic composition in Monte Carlo burnup calculations is presented to provide recommendations for time step scheme development in bRAPID. Two novel algorithms are implemented into bRAPID, which address: i) the generation of time-dependent correction factors for the fission density distribution in boundary nuclear fuel assemblies within a reactor core; ii) the calculation of pin-wise burnup distributions and isotopic concentrations.
Time step resolution analysis shows that a variable time step scheme, developed to accurately characterize important isotope evolution, can be used to optimize burnup calculations and minimize computation time. The two new algorithms have been benchmarked against the Monte Carlo code system Serpent. Results indicate that the time-dependent boundary correction algorithm improves fission density distribution calculations by including a more detailed representation of boundary physics. The pin-wise burnup algorithm expands bRAPID capabilities to provide material composition data at the pin level, with accuracy comparable to the reference calculation. In addition, wall-clock time analyses show that burnup calculations performed using bRAPID with these novel algorithms require a fraction of the time of Serpent. / Master of Science / Fuel burnup modeling is an important aspect of nuclear reactor design that provides information about the energy extracted (called burnup) and isotopes created or used in the fuel of a reactor over time. A reactor core is a collection of fuel assemblies, and assemblies are simply a bundle of fuel pins, which contain nuclear fuel such as Uranium. The desire for accurate and fast computer codes to calculate fuel burnup rises each year as engineers working in reactor core design seek to arrange fuel assemblies in an optimal pattern to extract the most energy. State of the art burnup codes exist, however they have certain limitations due to their underlying methodologies.
To satisfy this need, the bRAPID algorithm was developed by the Virginia Tech Transport Theory Group (VT³G). bRAPID is a new methodology capable of performing full core fuel burnup calculations in real time. bRAPID is able to calculate the criticality and burnup distribution of a reactor orders of magnitude faster than comparable algorithms, while addressing many of the shortcomings seen in other burnup codes.
In this thesis, studies of standard burnup codes are conducted in order to aid in bRAPID analysis: first in the form of a detailed study of the reference Monte Carlo model used in this thesis, and secondly in an investigation of the effect of time step selection – or the time intervals used in burnup calculations – on isotope concentration. Both of these studies are conducted using the benchmark code system, Serpent, with the latter study providing useful insight that can be used for bRAPID database development. This thesis then presents two new algorithms for bRAPID that expand its capability and improve performance. First, an algorithm to more accurately simulate the boundary regions of the core – called the time dependent boundary correction algorithm – is presented and benchmarked. Next, an algorithm to expand bRAPID capability from assembly-wise to pin-wise burnup calculations is implemented and tested. These two algorithms are benchmarked against the Serpent Monte Carlo based burnup code.
|
Page generated in 0.0126 seconds