SEARCH · Engineering Papers
Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.
Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.
This report investigates free expansion of Aluminum and provides a take home message of "The physically realistic SNAP machine-learning potential captures liquid-vapor coexistence behavior for free expansion of aluminum at a level not generally accessible to hydrocodes".
A common challenge with supervised machine learning (ML) in high energy physics (HEP) is the reliance on simulations for labeled data, which can often mismodel the underlying collision or detector response. To help mitigate this problem of domain shift, we propose RINO (Renormalization Group Invariance with No Labels), a self-supervised learning approach that can instead pretrain models directly on collision data, learning embeddings invariant to renormalization group flow scales. In this work, we pretrain a transformer-based model on jets originating from quantum chromodynamic (QCD) interactions from the JetClass dataset, emulating real QCD-dominated experimental data, and then finetune on the JetNet dataset -- emulating simulations -- for the task of identifying jets originating from top quark decays. RINO demonstrates improved generalization from the JetNet training data to JetClass data compared to supervised training on JetNet from scratch, demonstrating the potential for RINO pretraining on real collision data followed by fine-tuning on small, high-quality MC datasets, to improve the robustness of ML models in HEP.
Free-energy landscapes and chemical potentials govern the dynamics of phase transitions, transport, and stability in functional materials, yet they remain experimentally inaccessible under realistic operating conditions. Here we introduce a Bayesian model-integrated neural network (BMINN) that embeds physics-based formulations of non-autonomous partial differential-algebraic equations into probabilistic learning. This approach reconstructs hidden thermodynamics directly from macroscopic current-voltage data, providing quantitative access to metastable states, staging transitions, and energy barriers without synchrotron probes. Demonstrated on lithium-graphite electrodes, BMINN recovers full Gibbs free-energy landscapes with fidelity validated against operando X-ray diffraction. The framework generalizes across dynamical regimes, enabling accurate voltage prediction, internal state estimation, and inference of governing parameters. Beyond batteries, BMINN exemplifies a broadly applicable strategy for learning missing physics in multiphase, non-equilibrium systems, offering a new pathway to uncover hidden thermodynamic functions across condensed matter and materials physics.
This grant aimed to accelerate the development of specialized enzymes—biological catalysts essential for sustainable manufacturing and medicine—by integrating traditional laboratory evolution with cutting-edge artificial intelligence. To achieve this, we developed a suite of high-throughput sequencing tools and a centralized database to bridge the gap between a protein’s genetic "code" and its physical function. By training machine learning models on large datasets, we also demonstrated the ability to move beyond slow, trial-and-error testing to a "generative" approach, where AI can independently design new, versatile enzymes like tryptophan synthases. Ultimately, these findings demonstrate that combining laboratory data with computer-guided design enables the engineering of highly efficient biological tools with unprecedented speed and precision.
Megaampere-scale electrical wire experiments (EWEs) provide a platform for studying magnetohydrodynamic (MHD) instability growth in magneto-inertial fusion (MIF) devices. Even when nonlinear simulations of these experiments can digitally reproduce much of the experimentally observed instability growth, interpreting the results and understanding mode growth and evolution can be non-trivial. As a first step toward providing better interpretation of these simulation features, this work investigates the use of a deep neural network that uses Koopman operator theory to analyze the dynamics of pulsed-power-driven explosions of EWEs. This deep neural network is trained on 1-D resistive MHD simulations of EWEs. This neural network learns to transform the nonlinear data into a lower-dimensional representation where the time dynamics are linear. Layers of this neural network are shown to learn features of the simulations, including the locations of shock waves and different physical regimes of the simulation. Using the learned features, the network can compress a time state of the simulation consisting of 5120 data point into a 36-parameter lower-dimensional latent space embedding. Furthermore, these embeddings are shown to be clustered in the latent space by initial radius and time state.
Not Available
The original goal of this project was to diagnose and improve CLUBB, the turbulence and cloud fraction parameterization used in DOE’s E3SM model. For this purpose, we planned to use large-eddy simulation (LES) and ARM observations from the NE Pacific Ocean and the SGP site, in coordination with ARM’s LASSO program. It was determined that for these cloud regimes, many of the turbulent ‘moment closures’ which underlie CLUBB’s mathematical formulation are inconsistent with LES, which is an appropriate benchmark for testing this. The research assistant found that these closures could be more accurately formulated using machine learning using the LES as a training dataset. However the resulting parameterization proved to quickly drift away from physical plausibility. A novel machine-learning based boundary layer parameterization called MARBLE based on matching the time evolution of a parameterized cloud-topped boundary layer to reanalysis was then developed and published.
The Cyclotron Radiation Emission Spectroscopy (CRES) technique pioneered by Project 8 measures cyclotron radiation from individual electrons in a background magnetic field to construct a highly precise energy spectrum for beta decay studies and other applications. The detector, magnetic trap geometry, and electron dynamics give rise to a multitude of complex electron signal structures which carry information about distinguishing physical traits. Here, we develop machine learning models to classify CRES signals with high accuracy based on these traits, improve the resultant frequency spectrum, and offer the potential for a sophisticated analysis which will help Project 8 achieve tritium endpoint measurement in the future.
Stream temperature is a fundamental control on ecosystem health. Recent efforts incorporating process guidance into deep learning models for predicting stream temperature have been shown to outperform existing statistical and physical models. This performance is in part because deep learning architectures can actively learn spatiotemporal relationships that govern how water and energy propagate through a river network. However, exploration of how spatiotemporal awareness and process guidance influence a model's generalizability under shifting environmental conditions such as climate change is limited. Here, we use Explainable Artificial Intelligence (XAI) to interrogate how differing deep learning architectures affect a model's learned spatial and temporal dependencies, and how those learned dependencies affect a model's ability to maintain high accuracy when applied to unseen environmental conditions. Using the Delaware River Basin in the northeastern United States as a test case, we compare two spatiotemporally aware process–guided deep learning models for predicting stream temperature (a recurrent graph convolution network—RGCN, and a temporal convolution graph model—Graph WaveNet). Both models achieve equally high predictive performance when testing data are well represented in the training data (test root mean squared errors of 1.64°C and 1.65°C); however, Graph WaveNet significantly outperforms RGCN in 4 out of 5 experiments where test partitions represent different types of unseen environmental conditions. XAI results show that the architecture of Graph WaveNet leads to learned spatial relationships with greater fidelity to physical processes, and that this fidelity improves the generalizability of the model when applied to shifting and/or unseen environmental conditions.
Recent advances in quantum technologies have demonstrated that quantum systems can outperform classical ones in specific tasks, a concept known as quantum advantage. Although previous efforts have focused on computational speedups, a definitive and provable quantum advantage that is unattainable by any classical system has remained elusive. Here, in this work, we demonstrate a provable photonic quantum advantage by implementing a quantum-enhanced protocol for learning a high-dimensional physical process. Using imperfect Einstein–Podolsky–Rosen entanglement, we achieve a sample complexity reduction of 11.8 orders of magnitude compared to classical methods without entanglement. These results show that large-scale, provable quantum advantage is achievable with current photonic technology and represent a key step toward practical quantum-enhanced learning protocols in quantum metrology and machine learning.
Artificial intelligence and machine learning (AI/ML) are becoming important tools for scientific modeling and simulation as in several other fields such as image analysis and natural language processing. ML techniques can leverage the computing power available in modern systems and reduce the human effort needed to configure experiments, interpret and visualize results, draw conclusions from huge quantities of raw data, and build surrogates for physics based models. Domain scientists in fields like fluid dynamics, microelectronics and chemistry can automate many of their most difficult and repetitive tasks or improve the design times by use of the faster ML-surrogates. However, modern ML and traditional scientific highperformance computing (HPC) tend to use completely different software ecosystems. While ML frameworks like PyTorch and TensorFlow provide Python APIs, most HPC applications and libraries are written in C++. Direct interoperability between the two languages is possible but is tedious and error-prone. In this work, we show that a compiler-based approach can bridge the gap between ML frameworks and scientific software with less developer effort and better efficiency. We use the MLIR (multi-level intermediate representation) ecosystem to compile a pre-trained convolutional neural network (CNN) in PyTorch to freestanding C++ source code in the Kokkos programming model. Kokkos is a programming model widely used in HPC to write portable, shared-memory parallel code that can natively target a variety of CPU and GPU architectures. Our compiler-generated source code can be directly integrated into any Kokkosbased application with no dependencies on Python or cross-language interfaces.
The U.S. Department of Energy’s (DOE) Princeton Plasma Physics Laboratory (PPPL) is a collaborative national center for fusion energy science, basic sciences, and advanced technology. The Laboratory has three major missions: (1) to develop the scientific knowledge and advanced engineering to enable fusion to power the U.S. and the world; (2) to advance the science of nanoscale fabrication for technologies of tomorrow; and (3) to further the development of the scientific understanding of the plasma universe from laboratory to astrophysical scales. PPPL’s Laboratory Directed Research and Development (LDRD) program supports and encourages creativity and innovation and contributes to its long-term viability. New scientific and technical research areas emerge and are nurtured through the program. Furthermore, new capabilities are developed to enable the Laboratory to meet its and DOE’s missions. The program is used to systematically diversify the Laboratory’s programs and mission. In the last few years, the program has started projects in nanomaterial synthesis, microelectronics, advanced x-ray spectroscopy, high-energy-density physics, superconducting magnet technology, machine learning and artificial intelligence, 3D magnetic fields to optimize fusion plasmas, integration of permanent magnets with simple high-field magnets to reduce the cost of producing complex 3D magnetic fields, advanced computational methods for predictive understanding and control of fusion plasma, development of quantum computing algorithms for plasma physics, liquid metal plasma-facing components for fusion reactors, virtual engineering, and plasma-based space propulsion. The program is also the vehicle to recruit and train talented scientists and engineers with the new skills needed to perform the Laboratory’s mission. Many of the new hires through the program go on to become world-class scientists and engineers in their fields. This report provides descriptions and accomplishments of those LDRD projects that were completed during fiscal years 2018 through 2020.
Due to the strong integration of real-world physics, OT deception platforms must operate differently than traditional IT deceptions. For instance, turning off a valve will be detected downstream by other sensors because the flow will reduce and stop. Additionally, controllers and applications leverage data from sensors to send control commands to each other. A believable deception must be integrated with the system to project the effects of events. An attack will likely attempt to control the physical process in a negative manner. To make the attacker believe they are achieving their objective, it must predict the effects of these actions, to a reasonable degree. Our approach to simulating a model to generate realistic decoy behavior is explored including description of two approaches: a physics model-based approach and a data driven approach. The performance of two machine learning techniques are investigated in their ability to learn a good enough model of the physics of the system.
The next generation of advanced materials is tending toward increasingly complex compositions. Synthesizing precise composition is time-consuming and becomes exponentially demanding with increasing compositional complexity. An experienced human operator does significantly better than a novice but still struggles to consistently achieve precision when synthesis parameters are coupled. The time to optimize synthesis becomes a barrier to exploring scientifically and technologically exciting compositionally complex materials. This investigation demonstrates an active learning (AL) approach for optimizing physical vapor deposition synthesis of thin-film alloys with up to five principal elements. We compared AL-based on Gaussian process (GP) and random forest (RF) models. The best performing models were able to discover synthesis parameters for a target quinary alloy in 14 iterations. We also demonstrate the capability of these models to be used in transfer learning tasks. RF and GP models trained on lower dimensional systems (i.e., ternary, quarternary) show an immediate improvement in prediction accuracy compared to models trained only on quinary samples. Furthermore, samples that only share a few elements in common with the target composition can be used for model pre-training. We believe that such AL approaches can be widely adapted to significantly accelerate the exploration of compositionally complex materials.
The practice of collider physics typically involves the marginalization of multi-dimensional collider data to uni-dimensional observables relevant for some physics task. In many cases, such as classification or anomaly detection, the observable can be arbitrarily complicated, such as the output of a neural network. However, for precision measurements, the observable must correspond to something computable systematically beyond the level of current simulation tools. In this work, we demonstrate that precision-theory-compatible observable space exploration can be systematized by using neural simulation-based inference techniques from machine learning. We illustrate this approach by exploring the space of marginalizations of the energy 3-point correlator to optimize sensitivity to the top quark mass. We first learn the energy-weighted probability density from simulation, then search in the space of marginalizations for an optimal triangle shape. Although simulations and machine learning are used in the process of observable optimization, the output is an observable definition which can be then computed to high precision and compared directly to data without any memory of the computations which produced it. We find that the optimal marginalization is isosceles triangles on the sphere with a side ratio approximately $1 : 1 : \sqrt{2}$ (i.e. right triangles) within the set of marginalizations we consider.
Aqueous organic redox flow battery (AORFB) is a promising cost-competitive technology for large-scale energy storage. Among existing work, the dihydroxyphenazine (DHP)-based AORFB has demonstrated high energy density and low-capacity degradation in 10 cm$^2$ cells during lab tests. However, its commercial-scale performance in more complex environments remains unknown, posing a barrier to commercialization. To address this gap, this work presents a comprehensive performance evaluation of a 780 cm$^2$ DHP-based AORFB by combining a physics-based numerical model, machine learning (ML)-based surrogate models, and ML-derived sensitivity quantification. Specifically, we first select 12 key battery parameters that include 10 physicochemical and 2 operation quantities, then select 6 performance metrics that include energy efficiency (EE), discharging capacity, charging energy, and power losses due to concentration, activation, and ohmic over-potentials. With such selection, 12800 combinations of the 12 parameters are subsequently generated using the Latin Hypercube Sampling method. Such combinations, together with 38 pre-defined State of Charge, are then integrated to a validated AORFB model developed in COMSOL to compute the performance metrics. With both input parameters and performance metrics, 60 deep neural network (DNN) surrogate models are then trained to approximate the relationship between the 10 physicochemical quantities and 6 performance metrics at each flow rate and current density. Sensitivity scores are then calculated based on the DNN models. Two additional sensitivity analysis tools, i.e., MARS, and SHAP, are also used to cross-validate the sensitivity scores from the DNN. The results demonstrate that 1) the standard potential ranks first in controlling EE and charging energy, 2) the membrane conductivity is most critical for power loss and EE, and 3) specific area and reaction rate control activation power loss.
Here using a model heat engine, we show that neural-network-based reinforcement learning can identify thermodynamic trajectories of maximal efficiency. We consider both gradient and gradient-free reinforcement learning. We use an evolutionary learning algorithm to evolve a population of neural networks, subject to a directive to maximize the efficiency of a trajectory composed of a set of elementary thermodynamic processes; the resulting networks learn to carry out the maximally efficient Carnot, Stirling, or Otto cycles. When given an additional irreversible process, this evolutionary scheme learns a previously unknown thermodynamic cycle. Gradient-based reinforcement learning is able to learn the Stirling cycle, whereas an evolutionary approach achieves the optimal Carnot cycle. Our results show how the reinforcement learning strategies developed for game playing can be applied to solve physical problems conditioned upon path-extensive order parameters.
The Lee research group is a part of the Nuclear Lattice Effective Field Theory Collaboration, which has developed and performed ab initio lattice simulations of nuclear structure, scattering, reactions, and many-body systems. The other senior members of the collaboration include Ulf-G. Meißner at Bonn/Julich, Evgeny Epelbaum and Hermann Krebs at Bochum, Timo Lahde and Thomas Luu at Julich, and Gautam Rupak at Mississippi State. Our letter “Ab initio alpha-alpha scattering” was featured in a Nature News and Views article. Another letter “Nuclear binding near a quantum phase transition” was highlighted in a Viewpoint article in the online APS journal Physics as well as a news article in the IOP publication Physics World (September 21, 2016). Our letter “Eigenvector continuation with subspace learning” was also highlighted a Synopsis article in Physics. There are many promising ab initio approaches being used to calculate the properties of few-and many-nucleon systems. This includes the no-core shell model, symmetry-adapted no-core shell model quantum Monte Carlo, auxiliary-field diffusion Monte Carlo, self-consistent Green’s functions, many-body perturbation theory, in-medium similarity renormalization group, and coupled cluster methods. Lattice effective field theory is another ab initio approach which combines the framework of effective field theory with lattice Monte Carlo methods to allow favorable scaling from few- to many-body systems. Perhaps the most important aspect of lattice effective field theory is that its strengths and weaknesses are orthogonal to that of other ab initio methods. For example, lattice effective field theory has little difficulty in probing cluster structures and collectivity, while such features are much more difficult using other methods. Furthermore it can be used to compute superfluid condensate fractions as well as the phase diagram of strongly matter and the density and temperature dependence of clustering. Lattice effective field theory was first used in simulations of infinite nuclear matter and infinite neutron matter with pions and without pions. In addition to the efforts by our collaboration, there have been recent efforts by other groups as well.