Series
Master of Science in Nuclear Engineering
Master of Science in Nuclear Engineering
Permanent Link
Series Type
Degree Series
Description
Associated Organization(s)
Associated Organization(s)
Organizational Unit
119 results
Publication Search Results
Now showing
1 - 10 of 119
-
ItemMethodology for Multiphysics Validation of SILENE Reactor Experiments using MOOSE Tools(Georgia Institute of Technology, 2024-06-04) Grund, Nathan JohnMOOSE (Multiphysics Object-Oriented Simulation Environment) tools developed by Idaho National Laboratory are being utilized at increasing rates by the government, private companies, and universities alike. With this increase in use, validation of more diverse problems is important for showcasing the capabilities of the modeling software. A good example of this is the creation of a model for the SILENE reactor used in criticality transient experiments conducted in France. This system has unique feedback phenomena such as the creation of radiolytic gas that sets it apart from other types of reactors. In this work a framework for modeling the S3-300 SILENE experiment is assembled using Gmsh for the geometry, Serpent for multi-group cross section generation, and MOOSE tools for steady state, and later transient simulations. The thesis provides discussion on the background theory and calculations, as well as initial use of diffusion theory for solving the neutronics problem, before the eventual use of the point kinetics equations. The final method results in a model that can simulate the power and feedback, with solutions outlined to address lapses in accuracy due to the constraints of time and capabilities of the MOOSE code.
-
ItemModeling and Sensitivity Studies for Radioisotope Signatures of Molten Salt Reactors (MSRs)(Georgia Institute of Technology, 2024-04-29) Mitchell, Matthew JayDevelopments of MSR technologies are making rapid progress across the globe. MSRs are a type of Generation IV system chosen by the Gen IV International Forum (GIF) as being one of the most feasible Advanced Reactor Technologies (ART) for research and development. US companies have various planned designs driven by commercial interest and intend on licensing MSRs through the Nuclear Regulatory Commission (NRC). These reactor designs involve properly controlling radionuclides through off-gas systems. This work examines radioisotope emissions and assesses treaty detection technology to be used for monitoring the interference of human-made events. The research in this thesis analyzes various reactor activities with SCALE models on the nuclear forensic signals possible from a MSR in operation. Multiple Isotope Ratio Comparison (MIRC) plots are used to compare the activities of monitored radioisotopes of high interest in an MSR to the signatures of a highly enriched uranium (HEU) pulse. The sensitivity studies performed to determine the nuclear forensic signals of MSRs in operation is the main focus of this thesis. It is defined in detail the depletion analysis executed using the t6-depl sequence in TRITON and the point depletion and decay calculations using ORIGEN with a FLiBe fuel salt. MSR emissions from certain reactor applications may significantly overlap with signals produced in a HEU event and changing irradiation time, using hold tanks, and changing off-gas removal rates can significantly shift the radioisotope signatures away from a HEU pulse to avoid considerable global effects on the International Monitoring System (IMS).
-
ItemCategorization of additive manufacturing techniques for nuclear nonproliferation threat analysis(Georgia Institute of Technology, 2023-12-08) Cannon, Natalie L.Additive manufacturing, or AM, is a rapidly developing technology that simplifies and automates the production of intricate objects. Recently, AM methods have been imple- mented in the domains of nuclear weapons and nuclear enrichment technologies. However, there are presently limited international or domestic regulations for AM’s involvement in the nuclear sector, leading to unregulated proliferation pathways. Existing export regu- lations are broad in scope and do not account for the particular nuances of different AM techniques. It is crucial to scrutinize and assess the nuclear applications of AM methods to establish effective regulations and limitations for monitoring proliferation routes. This project involves identifying and assessing 31 of the most commonly employed AM meth- ods based on their potential impact on the nuclear fuel cycle. Using this identification and classification system, export controls can be directed at nuclear proliferation threats posed by AM, without disrupting the entire industry and fuel cycle. Additionally, this compre- hensive approach to regulating and monitoring proliferation channels would expose gaps in export regulations.
-
ItemCyber-resilience of machine learning-based digital twins for reactor autonomous control systems(Georgia Institute of Technology, 2023-11-03) Yockey, Patience ChristinaThe prevalence of Machine Learning (ML) technology in industrial control systems (ICS) and operational technology (OT) environments is on the rise. Autonomous control systems (ACS) based on ML digital twins (DTs) are being proposed as a new way to control nuclear power plant (NPP) subsystems using data-driven analytics and stochastic modeling. However, the implementation of ACS in advanced reactor controls must be assessed for its resilience against adversarial action, given the increasing incidence of cyber-attacks against critical infrastructure and OT environments. This thesis delves into the implementation techniques and cyber-resilience of ML-based ACS designs for advanced reactors. Its recommendations provide security and safeguards for physical and digital systems to protect against cyber-attacks on ML algorithms. Two implementations of ACS were considered: one as a virtual testbed using the International Atomic Energy Agency (IAEA) Asherah pressurized water reactor (PWR) nuclear power plant simulator, and one as a cyber-physical testbed using the Western Services Corporation (WSC) Generic Pressurized Water Reactor (GPWR) and real-time programmable logic controller (PLC) data. The purpose was to determine the impact of cyber-attacks on ACS in both a medium and high-fidelity environment. Cyber-attacks targeting ACS’ ML models’ training data, real-time data, and architecture were conducted to encapsulate potential targets against ML DTs and determine the impact and associated cyber risk of using ML-based DTs in control environments.
-
ItemA Transfer Function-Equivalent-Diffusion-based Approach for Monoenergetic Flux Prediction(Georgia Institute of Technology, 2023-10-31) Painter, Bailey ChristopherCoupled Monte Carlo (MC) and thermal hydraulic (TH) analysis is valuable as a design or reference tool but can be slow especially when implemented in a Picard iteration. Previous work has developed a novel prediction block to achieve convergence with fewer MC simulations. The prediction works in two stages: (1) a surrogate-like model predicts macroscopic cross sections on-the-fly and (2) a first order perturbation (FOP) solver predicts the flux response to the updated cross sections. The main challenge with the prediction block is that the FOP model requires a fission matrix. This fission matrix is impractical to generate for most problems, particularly fine-mesh problems. This work investigates replacing the FOP solution with a monoenergetic diffusion solution. For a simple 1D boiling water reactor (BWR) pincell, the proposed prediction block produces a monoenergetic flux that is within good agreement of the MC flux. A realistic 3D pressurized water reactor (PWR) core is then introduced to demonstrate that 3D, coarse mesh problems require the application of equivalence parameters such as assembly discontinuity factors (ADFs) and superhomogenization (SPH) factors to produce accurate 1-group nodal diffusion solutions. This work investigates the implementation of the well-established Jacobian-free Newton Krylov (JFNK) method to produce these equivalence parameters in a time-efficient manner. In this case, only tens of one-group nodal diffusion solutions are required to produce converged equivalence parameters. The results obtained in this work show that the converged equivalence parameters are very successful in reproducing the heterogenous solution (up to 2.5% error for SPH and 0.3% error for ADFs), without needing to modify the nodal diffusion solution. In addition, the results show that ADFs yield the best agreement and are also stable (i.e., not varying) when thermal hydraulic fields are perturbed. These results suggest that the proposed prediction block methodology is rigorous.
-
ItemModeling of Fluoride Molten Salt Reactor Depletion in SCALE(Georgia Institute of Technology, 2023-08-28) Bayne, ChristopherMolten Salt Reactors (MSRs) are Gen IV nuclear reactor designs that use fissile material dissolved in high-temperature molten salt. MSRs feature safety and economic benefits through their low operating pressure, combined fuel and coolant into one component, and high thermal efficiency. Due to the historically limited demand for modeling MSRs, there has been no established tool explicitly designed for such a task. The SCALE reactor modeling software suite, developed and maintained by Oak Ridge National Laboratory (ORNL), is selected as a tool to examine MSR modeling capabilities. SCALE has been widely used, verified, and benchmarked for modeling of light water reactors (LWRs) and it would prove beneficial to demonstrate and evaluate its use for MSRs. This thesis focuses on graphite-moderated, FLiBe-fueled MSRs. It aims to determine the impacts and trade-offs for a given SCALE MSR depletion model between the user-defined simulation parameters, accuracy of the depletion simulation results and the computational resources. Accuracy of the depletion simulations was determined through comparison of criticality estimates (keff) and isotopic compositions of key nuclides across the lifetime of the reactor simulation. The simulation parameters considered include: depletion sub-interval schemes (coarse vs fine depletion step schemes); trace-element tracking (addnux parameters 2 through 4), self-shielding methods (CENTRM vs BONAMI); cross-section libraries (ENDF/B-VII.1 252-multigroup vs continuous-energy); and specific power (10 MW/MTIHM vs 20 MW/MTIHM). All analyses were conducted up to burnups of 62 GWd/MTIHM. Additionally, the parallel performance on two supercomputing clusters (Idaho National Laboratory’s Sawtooth cluster and Georgia Institute of Technology’s PACE-Firebird cluster) was analyzed. Major findings include: • From coarser to finer depletion schemes. Maximum actinide percent differences experienced little effect (~ 0.32%), but percent difference of nuclides of significant radioactivity experienced significantly higher (~5.10%). Computational runtime was affected linearly with each additional depletion step introduced into the scheme. keff was statistically unaffected (<10 pcm difference). • From BONAMI to CENTRM self-shielding, runtime experienced relatively small change (6.81% increase), ~226 ppm average difference in keff, and ~2.51% maximum difference of actinides and nuclides of significant radioactivity. • From addnux=2 to addnux=4 nuclide tracking, runtime experienced only a 6.33% increase for MG. keff was significantly affected (167 pcm difference, growing to ~500 pcm at EOL). For CE addnux variation, the criticality and isotopic concentration differences were near indistinguishable from that of MG, but the computational runtime experienced a ~4.7x increase from addnux=2 to addnux=4. The difference between addnux=2 and addnux=3 was significantly smaller than that of addnux=3 to addnux=4, suggesting that addnux=4 contains nuclides with significantly more effect on reactivity for fluoride-fueled, graphite-moderated MSRs (potentially H-3 and He-3 stemming from larger Li-7 concentrations). • From MG to CE runtime experienced significant increase (~10x longer). Actinide maximum difference was ~7%, while the nuclides of significant radioactivity exhibited ~4% maximum difference. • Depletion at 20MW/MTIHM over 10 MW/MTIHM yielded differences that were expected due to the inherent changes in the reactor core physics, thus suggesting that analyses presented in this thesis can be scaled to other fluoride-fueled, graphite-moderated MSRs of different specific powers
-
ItemReference Multigroup Cross Section Results of Stylized Microreactor Benchmark Problems(Georgia Institute of Technology, 2022-12-09) Sommer, Aaron RyleyMicroreactors contain material and geometric features not characteristic to operating reactors. The usage of control material at the core periphery and the distance between absorber and fissionable assemblies allows for environmental effects to significantly affect multigroup data of a region given its boundary condition and location within the core. A microreactor design consisting of eighteen identical fuel assemblies surrounded by control drums rotated to determine reactivity is derived from literature with closed gaps in temperature data. This design is analyzed using the continuous-energy Monte Carlo code SERPENT 2 under six sets of environmental effects – four involving a single assembly with independently imposed radial and axial boundary conditions, and two involving a whole core model with drums rotated to minimize or maximize reactivity. The results generated include eigenvalues, pin power distributions, scattering kernels up to third order, and multigroup cross sections for capture, scattering, fission, and neutron production in selected regions in the problem sets. The dependence of these data on local environment imposed by a set of boundary conditions at assembly interfaces is investigated. The stylized benchmark model contains essential reactor physics qualities to significantly effect local multigroup cross sections of a given material at both differing axial or radial location within the core as well as core environment effects imposed by differing boundary conditions. These reference data can therefore benchmark computational methods and tools for multigroup cross section generation or core calculations based on pin power density.
-
ItemSurvey of Machine Learning Methods for Reactor Burnup Prediction(Georgia Institute of Technology, 2022-08-26) Cullen, Jonah NordstromThe development and spread of new nuclear technologies create a demand for new methods of safeguarding the fuel to prevent diversion for proliferation of nuclear weapons. One proposed method of controlled monitoring is the use of antineutrino detectors, which can be used to make assessments of isotopic inventory inside of the reactor based on the counts it detects externally. When continuous monitoring with such a detector is not available, the burnup of the reactor can be determined by placing the detector near the reactor, taking a measurement of antineutrino count rates across the energy spectrum, and then making burnup predictions from that given information. This thesis is a survey of machine learning methods that can be used to make such a prediction. The performance of each are compared to each other along with various feature engineering methods and hyperparameter selection to determine which model would be best for application in the field. Based on the studies performed in this survey, a simple ordinary least-squares polynomial regression of degree 3 and standard scaling is the fastest and most accurate method for predicting reactor burnup from antineutrino yield spectra when trained on similar labeled data. These results demonstrate that for small data set applications of this method, simple methods can outperform complex machine learning methods and should not be neglected in favor of something more complex.
-
ItemEvaluating system confidence of near-field antineutrino-based nuclear reactor safeguards(Georgia Institute of Technology, 2022-06-23) Dunbrack, MatthewThe International Atomic Energy Agency (IAEA) relies heavily on surveying facilities and verifying inventories to ensure that special nuclear material (SNM) pathways are correct and complete. This process, conducted through on-site inspections, draws a significant amount of the limited resources from the IAEA. Through implementing near-field antineutrino detection systems, changes in reactor core composition can be continuously monitored without the need of any expensive and invasive inspection. Our confidence in such a system, however, needs to be carefully considered for the IAEA to implement antineutrino detection systems for nuclear reactor safeguards. In this work, system confidence, or the certainty of the predicted antineutrino spectra, is evaluated to outline current antineutrino-based safeguard capabilities as well as to highlight the leading causes of uncertainty. The proposed system under evaluation is the Reactor Evaluation Through Inspection of Near-field Antineutrinos (RETINA) system, which utilizes high-fidelity modeling to predict the antineutrino spectra emitted from a simulated reactor. Certain deviations in real-time antineutrino spectra would indicate a shift in fissile inventory and a possible diversion of SNM from the reactor core. To fully analyze the role of reactor designs and diversion scenarios in the system evaluation, the antineutrino spectra was simulated for various next generation reactor designs as well as processed for possible diversion scenarios the IAEA would aim to detect. The results indicate that larger reactors with more common fissile inventories lead to lower system uncertainty. While some simulated diversion scenarios were consistently detected, the overlapping spectra led to low confidence of diversion following IAEA standards. Future work will go into modeling new reactor-detector systems as well as applying modern machine learning methods for confidence improvement.
-
ItemForensic Signatures from Laser Isotope Separation(Georgia Institute of Technology, 2022-05-04) Burns, Henry ScottAs part of its nonproliferation mission, the National Nuclear Security Administration (NNSA) notes, “Developing and maintaining the technical means to monitor whether the terms of a nuclear arms control treaty or other international agreement are fulfilled is a critical factor in ensuring that such agreements are successful.” As recognition grows that laser isotope separation (LIS) is commercially feasible, organizations such as the NNSA and IAEA are likely to require means of detecting and inspecting LIS facilities. It has long been recognized that LIS poses a proliferation risk, since the lower energy requirements and smaller physical parameters associated with its efficiency makes an LIS facility harder to detect. It is therefore necessary to determine novel ways of detecting the existence of an illicit LIS facility and confirming whether undeclared or safeguarded material was produced in an LIS facility. This work uses a mathematical model of LIS to determine the likely isotopic ratios in uranium enriched using an LIS process and provides a comparison to the isotopic ratios in centrifuge-enriched uranium. It also uses the model to determine several key operating parameters of an LIS device from the feed, tails, and product streams. This work is combined with an isotopic chronometer, and an analysis of hypothetical samples of uranium is given to demonstrate the qualities of an LIS facility that can be determined solely from the enriched and depleted uranium it produces.