Levi D. McClenny

Levi D. McClenny
Texas A&M University | TAMU · Department of Electrical and Computer Engineering

Ph.D. Electrical Engineering

About

13
Publications
2,905
Reads
How we measure 'reads'
A 'read' is counted each time someone views a publication summary (such as the title, abstract, and list of authors), clicks on a figure, or views or downloads the full-text. Learn more
86
Citations
Introduction
Levi McClenny holds a Ph.D. from Texas A&M University. His research centers around the intersection of deep learning and physics, and primary focuses include scientific machine learning and Physics-Informed Neural Networks

Publications

Publications (13)
Article
Uranium Dioxide (UO2) fuel powers almost all commercial Nuclear Power Plants (NPPs) worldwide, generating carbon-free energy and contributing to the fight against climate change. UO2 fuel incurs damage and fractures due to large thermal gradients that develop across the fuel pellet during normal and transient operating conditions. A comprehensive u...
Preprint
Full-text available
Physics-informed Neural Network (PINN) is a promising tool that has been applied in a variety of physical phenomena described by partial differential equations (PDE). However, it has been observed that PINNs are difficult to train in certain "stiff" problems, which include various nonlinear hyperbolic PDEs that display shocks in their solutions. Re...
Preprint
Full-text available
Physics-informed Neural Network (PINN) is a promising tool that has been applied in a variety of physical phenomena described by partial differential equations (PDE). However, it has been observed that PINNs are difficult to train in certain "stiff" problems, which include various nonlinear hyperbolic PDEs that display shocks in their solutions. Re...
Preprint
Full-text available
Commercial nuclear power plants extensively rely on fission energy from uranium dioxide (UO2) fuel pellets that provide thermal energy; consequently, generating carbon-free power in current generation reactors. UO2 fuel incurs damage and fractures during operation due to large thermal gradients that develop across the fuel pellet during normal oper...
Preprint
Full-text available
Physics-Informed Neural Networks promise to revolutionize science and engineering prac- tice, by introducing domain-aware deep machine learning models into scientific compu- tation. Several software suites have emerged to make the implementation and usage of these architectures available to the research and industry communities. Here we introduce T...
Preprint
Full-text available
Physics-Informed Neural Networks (PINNs) have emerged recently as a promising application of deep neural networks to the numerical solution of nonlinear partial differential equations (PDEs). However, it has been observed that the original PINN algorithm can produce inaccuracies around sharp transitions in the solution, as well as display instabili...
Preprint
Full-text available
In many real-world applications of deep learning, estimation of a target may rely on various types of input data modes, such as audio-video, image-text, etc. This task can be further complicated by a lack of sufficient data. Here we propose a Deep Multimodal Transfer-Learned Regressor (DMTL-R) for multimodal learning of image and feature data in a...
Article
Uncovering links between processing conditions, microstructure, and properties is a central tenet of materials analysis. It is well known that microstructure determines properties, but expressing these structural features in a universal quantitative fashion has proved to be extremely difficult. Recent efforts have focused on training supervised lea...
Article
Uncovering links between processing conditions, microstructure, and properties is a central tenet of materials analysis. It is well known that microstructure determines properties, but expressing these structural features in a universal quantitative fashion has proved to be extremely difficult. Recent efforts have focused on training supervised lea...
Article
Full-text available
Background Gene regulatory networks govern the function of key cellular processes, such as control of the cell cycle, response to stress, DNA repair mechanisms, and more. Boolean networks have been used successfully in modeling gene regulatory networks. In the Boolean network model, the transcriptional state of each gene is represented by 0 (inacti...
Conference Paper
Full-text available
This paper is concerned with optimal estimation of the state of a Boolean dynamical systems observed through correlated noisy Boolean measurements. The optimal Minimum Mean-Square Error (MMSE) state estimator for general Partially-Observed Boolean Dynamical Systems (POBDS) can be computed via the Boolean Kalman Filter (BKF). However, thus far in th...

Network

Cited By

Projects

Projects (3)
Project
In order to improve economy and efficiency of electricity production, the U.S. nuclear industry is considering increasing the fuel peak burnup (BU) beyond the current regulatory limit of 62 GWD/MTU. the Nuclear Regulatory Commission (NRC) will likely require nuclear power plants (NPPs) to analyze a number of potential operational occurrences and their potential consequences before such extension can be approved. A major factor in analyzing such scenarios is the behavior of fuel rods at high burnup (HBU). It is well-established that for LWR fuels, the fission gas release rate and probability of fuel fragmentation rapidly increase at HBU. This occurred across all reactor types, vendor supplied fuel and fuel cycle management. The underlying mechanisms by which these processes take place are still poorly understood. Nonetheless, there is a consensus that the drastic change of microstructure across the fuel pellet holds the key for understanding these mechanisms. From the high burnup structure (HBS) with large bubbles and fine grains close to the rim to the unrestructured hot central region or the transitional intermediate region in between, it is expected these distinct regions play important roles at different steady and transient conditions. Here, we employ a multi-physics mesoscale model to estimate the rates of gas bubble swelling/release and fracture. Specifically, we construct a model that couples rate-theory, phase-field, and finite-element methods to account for the production and clustering of defects, the formation of sub-grains, and the resultant change in fracture toughness. The effects of defect production and migration rates, the underlying microstructure, and temperature are thoroughly investigated. Parametric studies and sensitivity analysis are used to asses uncertainty.
Project
The BoolFilter package provides tools for state estimation and inference of partially-observed Boolean dynamical systems. This package contains functions for simulating data, obtaining the optimal estimator in the presence of various observation models, applying the optimal MMSE filter and smoother, particle filter implementation of optimal filter, as well as means of parameter estimation and network inference. The package BoolFilter is freely available from the R project at https://cran.r-project.org/web/packages/BoolFilter/index.html