arXiv subjects

Michael J. Mueterthies

Publications and source records attributed to Michael J. Mueterthies.

5 recordsLinked to original sources

arXiv2026

A Methodology for Thermal Limit Bias Predictability Through Artificial Intelligence

Nuclear power plant operators face significant challenges due to unpredictable deviations between offline and online thermal limits, a phenomenon known as thermal limit bias, which leads to conservative design margins, increased fuel costs, and operational inefficiencies. This work presents a deep learning based methodology to predict and correct this bias for Boiling Water Reactors (BWRs), focusing on the Maximum Fraction of Limiting Power Density (MFLPD) metric used to track the Linear Heat Generation Rate (LHGR) limit. The proposed model employs a fully convolutional encoder decoder architecture, incorporating a feature fusion network to predict corrected MFLPD values closer to online measurements. Evaluated across five independent fuel cycles, the model reduces the mean nodal array error by 74 percent, the mean absolute deviation in limiting values by 72 percent, and the maximum bias by 52 percent compared to offline methods. These results demonstrate the model's potential to meaningfully improve fuel cycle economics and operational planning, and a commercial variant has been deployed at multiple operating BWRs.

cs.LG↗

arXiv2025

Optimization through In-Context Learning and Iterative LLM Prompting for Nuclear Engineering Design Problems

The optimization of nuclear engineering designs, such as nuclear fuel assembly configurations, involves managing competing objectives like reactivity control and power distribution. This study explores the use of Optimization by Prompting, an iterative approach utilizing large language models (LLMs), to address these challenges. The method is straightforward to implement, requiring no hyperparameter tuning or complex mathematical formulations. Optimization problems can be described in plain English, with only an evaluator and a parsing script needed for execution. The in-context learning capabilities of LLMs enable them to understand problem nuances, therefore, they have the potential to surpass traditional metaheuristic optimization methods. This study demonstrates the application of LLMs as optimizers to Boiling Water Reactor (BWR) fuel lattice design, showing the capability of commercial LLMs to achieve superior optimization results compared to traditional methods.

cs.LG↗

arXiv2021

Some Recent Developments in 5th Force Searches

Recently we drew attention to the fact that most recent 5th force searches and tests of the weak equivalence principle (WEP) utilize only one or two pairs of test samples. We argue that, despite the great precision of these experiments, the lack of diversity of samples may mean they are unable to detect new composition-dependent forces, which requires the observation of a pattern in the data. Such a pattern was observed in the experiment by Eötvös, Pekár, and Fekete (EPF), the last high precision test of the WEP which used a significant number of different samples. We advocate for new experiments utilizing a sufficient number and range of samples to either confirm or refute the pattern found in the EPF experimental data.

hep-ph↗

arXiv2021

Indications of a Fifth Force Coupling to Baryon Number in the Potter Test of the Weak Equivalence Principle

We have reanalyzed data obtained by Potter in a 1923 experiment aimed at testing whether the accelerations of test masses in the Earth's gravitational field are independent of their compositions. Although Potter concludes that the accelerations of his samples compared to a brass standard were individually consistent with a null result, we show that the pattern formed from a combined plot of all of his data suggests the presence of a fifth force coupling to baryon number.

hep-ph↗

arXiv2021

Significance of Composition-Dependent Effects in Fifth-Force Searches

Indications of a possible composition-dependent fifth force, based on a reanalysis of the Eötvös experiment, have not been supported by a number of modern experiments. Here, we argue that searching for a composition-dependent fifth force necessarily requires data from experiments in which the acceleration differences of three or more independent pairs of test samples of varying composition are determined. We suggest that a new round of fifth-force experiments is called for, in each of which three or more different pairs of samples are compared.

hep-ph↗