Physical Sciences

Recent works have implemented machine learning based solutions for many complex classification tasks including pulse shape discrimination in radiation detection. The present work aims to advance the application of machine learning to pulse shape discrimination in neutron detection. A machine learning based neutron-gamma discrimination technique is investigated for various neutron energy distributions produced from DD, DT, (α,n), and spontaneous fission neutron sources. Comprehensive investigations on the training data generation techniques, the impact of the PMT bias, and the discrimination performance are conducted. With the increase of the PMT bias voltage, the neutron classification performance peaked at 1500 V with 81 % of validation neutrons being identified at a false positive rate of 1E-6 while the further bias increase led to a notable degradation in performance. The unsatisfactory classification performance encountered when training off of one neutron source type and classifying neutrons from the other source types was greatly improved with the application of the transfer learning techniques. The remaining variation in the performance was accounted for by the energy dependence of the neutron classification. It was demonstrated that at the 1E-6 FPR specificity level, the events within the region of overlap for neutron and photon populations could be separated, down to a detected energy of 30 keVee. An overall intrinsic neutron detection efficiency of 12.5 % was achieved for the ²⁵²Cf neutron source at a false positive rate of 1E-6.

Abstract: Future, high-fidelity simulations of multi-GeV-class Laser Wakefield Accelerators (LWFAs) will need to model the propagation of high-intensity laser drivers over meter-scale plasmas with high spatial and temporal resolutions, thus requiring high amounts of computational resources. Various techniques have been devised over the years to reduce the computational cost of such simulations, including the time-averaged ponderomotive approximation, and the use of the Lorentz boosted frame technique. In this paper we discuss the combination of these two computational techniques, highlighting the resulting significant reduction in the computational cost of LWFA simulations and the limitations of this approach. The combination of the two techniques can potentially become essential for the modeling of a multi-TeV, LWFA-based collider.

Efficient accelerator modeling and particle tracking are key for the design and configuration of modern particle accelerators. In this work, we present JuTrack, a nested accelerator modeling package developed in the Julia programming language and enhanced with compiler-level automatic differentiation (AD). With the aid of AD, JuTrack enables rapid derivative calculations in accelerator modeling, facilitating sensitivity analyses and optimization tasks. We demonstrate the effectiveness of AD-derived derivatives through several practical applications, including sensitivity analysis of space-charge-induced emittance growth, nonlinear beam dynamics analysis for a synchrotron light source, and lattice parameter tuning of the future Electron-Ion Collider (EIC). Through the incorporation of automatic differentiation, this package opens up new possibilities for accelerator physicists in beam physics studies and accelerator design optimization. Program summary: Program Title: JuTrack CPC Library link to program files: https://doi.org/10.17632/r2g5zkwp7s.1 Developer's repository link: https://github.com/MSU-Beam-Dynamics/JuTrack.jl.git Licensing provisions: MIT Programming language: Julia Nature of problem: Derivatives of the physics parameters calculated in accelerator modeling are critical for sensitivity analysis and optimization of the whole system. Traditional numerical approaches often rely on finite differences for derivative computations, which can prone to numerical inaccuracies. In highly nonlinear accelerator systems, like those encountered in synchrotrons and colliders, accurate sensitivity analysis and optimization require a large number of derivative evaluations. Thus, there is a need for more efficient methods to compute these derivatives accurately, especially when optimizing complex accelerator lattices or studying complicated collective effects, such as space-charge effects, wakefield effects, and beam-beam interaction. Solution method: JuTrack addresses this problem by integrating compiler-level automatic differentiation (AD) into accelerator modeling routines, offering a powerful toolset for rapid derivative computation. Developed in the Julia programming language, JuTrack uses the Enzyme AD package to perform gradient-based analyses with minimal computational overhead. The package provides an efficient way to compute derivatives by directly differentiating through the model code, thus avoiding approximation errors associated with finite difference methods. It is designed to handle complex beam dynamics simulations, including complicated collective effects, such as space-charge effects, wakefield effects, beam-beam interaction, and combination of Truncated Power Series Algebra (TPSA) with AD. It can be applied to lattice optimization and beam dynamics analysis for future accelerators like the Electron-Ion Collider (EIC). Users can easily apply the package to their models, enabling robust optimization and sensitivity analysis in their accelerator studies. Additional comments including restrictions and unusual features: JuTrack is particularly well-suited for scenarios requiring frequent derivative calculations, such as during beam dynamics optimization, sensitivity analysis, and accelerator tuning. Its integration with the Julia programming language provides excellent performance due to Julia's just-in-time (JIT) compilation capabilities. The modular nature of JuTrack and Julia's easy-to-understand syntax allows for future extensions and custom modifications, making it adaptable to a variety of accelerator configurations.

Accelerator physics presents unique challenges for natural language processing (NLP) due to its specialized terminology and complex concepts. A key component in overcoming these challenges is the development of robust text embedding models that transform textual data into dense vector representations, facilitating efficient information retrieval and semantic understanding. In this work, we introduce AccPhysBERT, a sentence embedding model fine-tuned specifically for accelerator physics. Our model demonstrates superior performance across a range of downstream NLP tasks, surpassing existing models in capturing the domain-specific nuances of the field. We further showcase its practical applications, including semantic paper-reviewer matching and integration into retrieval-augmented generation systems, highlighting its potential to enhance information retrieval and knowledge discovery in accelerator physics. Published by the American Physical Society 2025

Modern laser plasma accelerators often require plasma waveguides tens of centimeters long to propagate a high-intensity drive laser pulse. Tapering the longitudinal gas density profile in 10 cm scale gas jets could allow for single stage laser plasma acceleration well beyond 10 GeV with current petawatt-class laser systems. Via simulation and interferometry measurements, we show density control by longitudinally adjusting the throat width and jet angle. Density profiles appropriate for tapering were calculated analytically and via particle-in-cell simulations and were matched experimentally. These simulations show that tapering can increase electron beam energy using 19 J laser energy from ∼9 GeV to >12 GeV in a 30 cm plasma and the accelerated charge by an order of magnitude.

Intermetallic phases are preferred to reduce the amount of platinum used for catalytic applications as compared to solid solution alloys, due to their stability at elevated temperatures while preserving or even enhancing the catalytic properties. Here, we show a two-step process to form an intermetallic NiPt L10 phase. In this work, NiPt solid solution thin films were fabricated by direct current and high-power impulse magnetron sputtering processes, which allow for precise thickness and chemical composition control. Following deposition, an additional annealing step is used to form the desired intermetallic phase. We show that the required annealing time for intermetallic phase formation is considerably reduced for NiPt thin films with a thickness of 240 nm, as compared to its bulk counterpart.

Residual resistance ratio (RRR) of superconducting strands is an important parameter for magnet electrical stability. RRR serves as a measure of the low-temperature electrical conductivity of the copper within a conductor that has a copper stabilization matrix. For Nb3Sn, due to the need of a reaction heat treatment, the technical requirements for high quality measurements of strands extracted from Rutherford cables are particularly demanding. Quality of wire, cabling deformation, heat treatment temperature, heat treatment atmosphere, sample handling, and measurement methods can all affect the RRR. Therefore, as an integral part of the electrical quality control (QC) of Nb3Sn Rutherford cables manufactured at the Lawrence Berkeley National Laboratory, it was prudent that we established a RRR measurement system that can isolate the assessment of cable-fabrication-related impacts from sample preparation and measurement factors. Here we describe a bespoke cryocooler-based measurement system, capable of measuring RRR of over 80 samples in a single cooldown. The samples are mounted on custom-designed printed circuit boards that accommodate the shape of strands extracted from a Rutherford cable without added deformation, which we will show is critical in ensuring that the measurements accurately represent the RRR values of the conductor within the cable. Using this sample mounting solution, we routinely measure the overall RRR of the strand as well as individual intra-strand sections corresponding to both cable edges and cable broad faces with high reproducibility. Such measurements provide valuable information on the variation of RRR along the length of the strands as well as across strand productions and cable runs over time.