Search NASA⌕ Search

SEARCH · Search NASA

Results for “differentiable physics”

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.

At least 145 records · Page 8
All sourcesNASA NTRSDOE OSTI
DOE OSTI2024

Exact enforcement of temporal continuity in sequential physics-informed neural networks

The use of deep learning methods in scientific computing represents a potential paradigm shift in engineering problem solving. One of the most prominent developments is Physics-Informed Neural Networks (PINNs), in which neural networks are trained to satisfy partial differential equations (PDEs). While this method shows promise, the standard version has been shown to struggle in accurately predicting the dynamic behavior of time-dependent problems. To address this challenge, methods have been proposed that decompose the time domain into multiple segments, employing a distinct neural network in each segment and directly incorporating continuity between them in the loss function of the minimization problem. In this work we introduce a method to exactly enforce continuity between successive time segments via a solution ansatz. This hard constrained sequential PINN (HCS-PINN) method is simple to implement and eliminates the need for any loss terms associated with temporal continuity. The method is tested for a number of benchmark problems involving both linear and non-linear PDEs. Examples include various first order time dependent problems in which traditional PINNs struggle, namely advection, Allen–Cahn, and Korteweg–de Vries equations. Furthermore, second and third order time-dependent problems are demonstrated via wave and Jerky dynamics examples, respectively. Notably, the Jerky dynamics problem is chaotic, making the problem especially sensitive to temporal accuracy. Finally, the numerical experiments conducted with the proposed method demonstrated superior convergence and accuracy over both traditional PINNs and the soft-constrained counterparts.

42 ENGINEERING↗
DOE OSTI2025

First Differential Measurement of the Single 𝜋 + Production Cross Section in Neutrino Neutral-Current Scattering

Since its first observation in the 1970s, neutrino-induced neutral-current single positive pion production (NC⁢1⁢𝜋 + ) has remained an elusive and poorly understood interaction channel. This process is a significant background in neutrino oscillation experiments and studying it further is critical for the physics program of next-generation accelerator-based neutrino oscillation experiments. In this Letter, we present the first double-differential cross-section measurement of NC⁢1⁢𝜋 + interactions using data from the ND280 detector of the T2K experiment collected in 𝜈-beam mode. The measured flux-averaged integrated cross section is 𝜎 = (6.07 ± 1.22) × 10 −41 cm 2 /nucleon. We compare the results on a hydrocarbon target to the predictions of several neutrino interaction generators and final-state-interaction models. While model predictions agree with the differential results, the data show a weak preference for a cross-section normalization approximately 30% higher than predicted by most models studied in this Letter.

Neutrino detection↗
DOE OSTI2025

First Measurement of $𝜈_𝑒$ and $\bar{𝜈}_𝑒$ Charged-Current Single Charged-Pion Production Differential Cross Sections on Argon Using the MicroBooNE Detector

Understanding electron neutrino interactions is crucial for measurements of neutrino oscillations and searches for new physics in neutrino experiments. We present the first measurement of the flux-averaged $𝜈_𝑒$ + $\bar{𝜈}_𝑒$ charged-current single charged-pion production cross section on argon using the MicroBooNE detector and data from the NuMI neutrino beam. The total cross section is measured to be [0.93 ± 0.13⁢(stat) ± 0.27⁢(syst)] × 10 −39 cm 2 /nucleon at a mean $𝜈_𝑒$ + $\bar{𝜈}_𝑒$ energy of 730 MeV. Differential cross sections are also reported in electron energy, electron and pion angles, and electron-pion opening angle.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2024

Differentiable modeling and optimization of non-aqueous Li-based battery electrolyte solutions using geometric deep learning

Electrolytes play a critical role in designing next-generation battery systems, by allowing efficient ion transfer, preventing charge transfer, and stabilizing electrode-electrolyte interfaces. In this work, we develop a differentiable geometric deep learning (GDL) model for chemical mixtures, DiffMix, which is applied in guiding robotic experimentation and optimization towards fast charging battery electrolytes. In particular, we extend mixture thermodynamic and transport laws by creating GDL-learnable physical coefficients. We evaluate our model with mixture thermodynamics and ion transport properties, where we show improved prediction accuracy and model robustness of Diff-Mix than its purely data-driven variants. Furthermore, with a robotic experimentation setup, Clio, we improve ionic conductivity of electrolytes by over 18.8% within 10 experimental steps, via differentiable optimization built on DiffMix gradients. By combining GDL, mixture physics laws, and robotic experimentation, DiffMix expands the predictive modeling methods for chemical mixtures and enables efficient optimization in large chemical spaces.

25 - ENERGY STORAGE↗
DOE OSTI2026

Massive right-handed neutrinos in $\bar{B} \to D^* τ\bar X$ decay

We explore signatures of a massive right-handed neutrino (RHN) in angular distributions of $\bar{B} \to D^* (\to D π) τ(\to πν_τ) \bar X$ decays, where $X$ is an invisible state. We assume the new physics is described by the standard model effective field theory extended with an RHN in the MeV-GeV mass range. We calculate for the first time the full differential distributions in terms of the visible final states, including the decay of the $τ$ lepton. We evaluate the sensitivity of various distributions to the new physics operators.

FOS: Physical sciences↗
DOE OSTI2024

Modeling In-Reactor Temperature Gradients in Lithium Aluminate Ceramic Pellets

In reactor thermal performance of lithium aluminate ceramic pellets are calculated to determine operating temperatures if internal components shift during irradiation. A 2D conduction model was setup to model this scenario with additional physics, including thermal expansion of components to reactor temperatures, creep down of the cladding due to differential pressure between the inside of the rod and reactor coolant system, and net radiation exchange between components. The non-linear coupling of the governing equations presented convergence challenges due to the very different effects from a small temperature change on the radiation vs radial and circumferential conduction equations. Initial temperature guesses for the multidimensional solution were provided from conservative 1D models simulating bounding scenarios. Convergence was greatly improved by taking small steps towards the root solutions and averaging with initial guesses. The final method robustly converged all 132 points at an axial plane after a few dozen iterations. The temperature results show that the additional physics have a secondary effect as the dominant heat transfer mechanism is radial conduction across the gas gap followed by circumferential conduction around components. Circumferential conduction in the lithium aluminate pellets is enhanced by the relatively high thermal conductivity of this ceramic.

Carstens, Nathan A.↗
DOE OSTI2025

An adaptive and stability-promoting layerwise training approach for sparse deep neural network architecture

This work presents a two-stage adaptive framework for progressively developing deep neural network (DNN) architectures that generalize well for a given training data set. In the first stage, a layerwise training approach is adopted where a new layer is added each time and trained independently by freezing parameters in the previous layers. We impose desirable structures on the DNN by employing manifold regularization, sparsity regularization, and physics-informed terms. We introduce a ε – δ – stability-promoting concept as a desirable property for a learning algorithm and show that employing manifold regularization yields a ε – δ stability-promoting algorithm. Further, we also derive the necessary conditions for the trainability of a newly added layer and investigate the training saturation problem. In the second stage of the algorithm (post-processing), a sequence of shallow networks is employed to extract information from the residual produced in the first stage, thereby improving the prediction accuracy. Numerical investigations on prototype regression and classification problems demonstrate that the proposed approach can outperform fully connected DNNs of the same size. Moreover, by equipping the physics-informed neural network (PINN) with the proposed adaptive architecture strategy to solve partial differential equations, we numerically show that adaptive PINNs not only are superior to standard PINNs but also produce interpretable hidden layers with provable stability. As a result, we also apply our architecture design strategy to solve inverse problems governed by elliptic partial differential equations.

42 ENGINEERING↗
DOE OSTI2024

Uncovering Backbone Conformation for Rigid DPP-Based Donor–Acceptor Conjugated Polymer Using Deuterium Labeling and Neutron Scattering

The conjugated polymer’s backbone conformation dictates the delocalization of electrons, ultimately affecting its optoelectronic properties. Most conjugated polymers can be viewed as semirigid rods with their backbone embedded among long alkyl side chains. Thus, it is challenging to experimentally quantify the conformation of a conjugated backbone. Here, we performed contrast variation neutron scattering on rigid conjugated donor–acceptor (D–A) diketopyrrolopyrrole (DPP) polymers with selectively deuterated side chains to measure the conjugated backbone conformation. We first synthesized DPPbased polymers with deuterated side chains, confirmed by NMR and FTIR. Using contrast variation neutron scattering, we found that the DPP-based conjugated polymers are much more rigid than poly(3-alkylthiophenes), with persistence length (L p ) at 16–18 nm versus 2–3 nm. More importantly, in contrast to the relatively flexible poly(3-alkylthiophenes) whose backbone is more flexible than the whole polymer, we found that the backbone of DPP-based polymers has the same L p value compared to the whole polymer chain. This indicates that side chain interference on backbone conformation is not present for the semirigid polymer, which is further confirmed by coarse-grained molecular dynamics (CG-MD) simulations. Our work provides a novel protocol to probe polymer’s backbone conformation and paradigm-shifting understanding of the backbone conformation of semirigid conjugated polymers.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2024

Theoretical investigation of decoherence channels in athermal phonon sensors

The creation and evolution of nonequilibrium phonons is central in applications ranging from cosmological particle searches to decoherence processes in qubits. However, the fundamental understanding of decoherence pathways for athermal phonon distributions in solid-state systems remains an open question. Using first-principles calculations, we investigate the primary decay channels of athermal phonons in two technologically relevant semiconductors—Si and GaAs. We quantify the contributions of anharmonic, isotopic, and interfacial scattering in these materials. From this, we construct a model to estimate the thermal power in a readout scheme as a function of time. We discuss the implication of our results on noise limitations in current phonon sensor designs and strategies for improving coherence in next-generation phonon sensors.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗
DOE OSTI2026

Neural posterior unfolding

Differential cross section measurements are the currency of scientific exchange in particle and nuclear physics. A key challenge for these analyses is the correction for detector distortions, known as deconvolution or unfolding. Binned unfolding of cross section measurements traditionally rely on the regularized inversion of the response matrix that represents the detector response, mapping pre-detector (`particle level') observables to post-detector (`detector level') observables. In this paper we introduce Neural Posterior Unfolding, a modern, Bayesian approach that leverages normalizing flows for unfolding. By using normalizing flows for neural posterior estimation, NPU offers several key advantages including implicit regularization through the neural network architecture, fast amortized inference that eliminates the need for repeated retraining, and direct access to the full uncertainty in the unfolded result. In addition to introducing NPU, we implement a classical Bayesian unfolding method called Fully Bayesian Unfolding (FBU) in modern Python so it can also be studied. These tools are validated on simple Gaussian examples and then tested on simulated jet substructure examples from the Large Hadron Collider (LHC). We find that the Bayesian methods are effective and worth additional development to be analysis ready for cross section measurements at the LHC and beyond.

Analysis and statistical methods↗
DOE OSTI2025

An Educational Guide for 2D Stellar Structure Calculations of Rapidly Rotating Stars using the ESTER code

The Evolution STEllaire en Rotation (ESTER) code is the first 2D stellar structure code to be made open-source and freely available to the astronomy and astrophysics community. An important and novel advancement of this code is that it can reproduce the distorted shape and observable signatures (e.g., gravity darkening) of rapidly rotating stars. ESTER also calculates the steady-state large-scale flows within the star, namely their differential rotation and associated meridional circulation. In this report, we explore and document the physics implemented within version 1.1.0rc2 of the ESTER code, in a way that complements published descriptions. We illustrate this physics by plotting how stellar structure parameters vary through stellar interiors at a range of latitudes and at different angular velocities. We investigate how the thin convective envelopes of intermediate mass stars vary with latitude when rapidly rotating, becoming deeper and thicker near the equator. Simple comparisons of ESTER model predictions (e.g., central temperature and density, luminosity) with the output from the Modules for Experiments in Stellar Astrophysics (MESA) code [Paxton et al., 2010] shows generally good agreement. Additional comparisons provide important benchmarking and verification for ESTER as a comparatively young code. Finally, we provide a guide for installing and running the code on our local university cluster, aimed at helping students to begin work.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2026

E-PINNs: Epistemic Physics-Informed Neural Networks

Physics-informed neural networks (PINNs) have demonstrated promise as a framework for solving forward and inverse problems involving partial differential equations. Despite recent progress in the field, it remains challenging to quantify uncertainty in these networks. While techniques such as Bayesian PINNs (B-PINNs) provide a principled approach to capturing epistemic uncertainty through Bayesian inference, they can be computationally expensive for large-scale applications. In this work, we propose Epistemic Physics-Informed Neural Networks (E-PINNs), a framework that uses a small network, the epinet, to efficiently quantify epistemic uncertainty in PINNs. The proposed approach works as an add-on to existing, pre-trained PINNs with a small computational overhead. We demonstrate the applicability of the proposed framework in various test cases and compare the results with B-PINNs using Hamiltonian Monte Carlo (HMC) posterior estimation and dropout-equipped PINNs (Dropout-PINNs). In our experiments, E-PINNs achieve calibrated coverage with competitive sharpness at substantially lower cost. We demonstrate that when B-PINNs produce narrower bands, they under-cover in our tests. E-PINNs also show better calibration than Dropout-PINNs in these examples, indicating a favorable accuracy-efficiency trade-off.

AI for Science↗
DOE OSTI2025

Order conditions for nonlinearly partitioned Runge-Kutta methods

Recently, a new class of nonlinearly partitioned Runge–Kutta (NPRK) methods was proposed for nonlinearly partitioned systems of autonomous ordinary differential equations y' = F(y, y). The target class of problems are those in which different scales, stiffnesses, or physics are coupled in a nonlinear way, wherein the desired partition cannot be written in a classical additive or component-wise fashion. Here we use a rooted-tree analysis to derive full-order conditions for NPRKM methods, where M denotes the number of nonlinear partitions. Due to the nonlinear coupling and thereby the mixed product differentials, it turns out that the standard node-colored rooted tree analysis used in analyzing ODE integrators does not naturally apply. Instead we develop a new edge-colored rooted-tree framework to address the nonlinear coupling. The resulting order conditions are enumerated, are provided directly for up to fourth order with M = 2 and third order with M = 3, and are related to existing order conditions of additive and partitioned RK methods. We conclude with an example that shows how the nonlinear order conditions can be used to obtain an embedded estimate of the state-dependent nonlinear coupling strength in a dynamical system.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2025

Understanding plasma turbulence through exact coherent structures

Plasma turbulence is a key challenge in understanding transport phenomena in magnetically confined plasmas. This work presents a generalized framework to analyze plasma turbulence that utilizes periodic orbit theory. In periodic orbit theory, doubly periodic solutions (coherent structures) of the governing equation(s) serve as building blocks of the considered turbulent dynamics. To illustrate the concept and method, the particularly simple Kuramoto–Sivashinsky (referred to here as LMRT for the original authors: LaQuey, Mahajan, Rutherford, and Tang) trapped-ion mode toy model is used. By applying numerical optimization techniques to the LMRT equation, we extract coherent spacetime patterns that represent the library of allowable fundamental structures of the equation. These structures provide a framework to systematically describe turbulence as a composition of recurrent solutions, revealing an underlying order within chaotic plasma motion. Although illustrated here using the simplified LMRT model for clarity, this framework provides a general strategy that can be extended to more complex and realistic models of plasma turbulence, including gyrokinetic systems. This offers a new method for predicting and potentially controlling transport processes in fusion plasmas by providing a bridge between nonlinear dynamical systems theory and plasma physics in the form of a generalized framework with which to analyze and understand spatially extended nonlinear partial differential equations.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2026

TTDAQ: A Continuous Flow, Timing and Trigger DAQ System

Final Scientific/Technical Report for DOE Award DE-SC0019581, “TTDAQ: A Continuous Flow, Timing and Trigger DAQ System.” The report summarizes Telluric Labs’ Phase II STTR work developing silicon-photonic building blocks for a software-defined, continuous-flow, trigger-less data acquisition system for next-generation high-energy and nuclear-physics detectors. The project focused on radiation-hard photonic integrated circuits, remote optical illumination, dense wavelength-division multiplexing, and a differential microring-resonator transceiver architecture designed to improve high-speed optical link stability and bandwidth. The report describes project objectives, technical accomplishments, AIM Photonics tape-outs, bench characterization, radiation-hardness testing, deferred integration work, and potential applications beyond physics readout.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
DOE OSTI2026

Convergence Criteria for Multiphysics Simulations

The behavior of engineered systems is often influenced by multiple physical phenomena, such as mechanical deformation, heat transfer, and chemical species transport and reactions. There are often strong interactions between these phenomena, and there is increasing interest in applying coupled-physics models to improve understanding of physical behavior under complex environmental conditions. Multiple simulation frameworks that facilitate coupled-physics simulations are in widespread use, and these employ a variety of techniques to account for interactions between those physics. Many frameworks solve the physics models independently and transfer results between them. Alternatively, a single monolithic system of equations for every physics model can be formed and solved. Each of these approaches has its benefits and drawbacks, and the optimal approach varies depending on the nature of the problem. The open-source MOOSE framework was developed targeting solution of large-scale multiphysics problems. Although it provides options for all these coupling approaches, its standard approach for multiphysics solutions is to form and solve a single monolithic system of equations containing the unknowns for all physics models. MOOSE provides a streamlined approach for users to define the solution variables, the terms in the partial differential equations pertaining to each variable, and interactions between solution variables. One aspect of the monolithic solution approach that can be problematic, however, is defining appropriate convergence criteria for the nonlinear system. A standard approach is to determine convergence is to simply take a norm of the residual vector corresponding to the full vector of unknowns. However, if the residual vector contains variables for multiple physics models, the magnitudes of those variables can differ significantly, and the variables can converge at significantly different rates from each other. It is important to ensure that the variables for each of the physics are converged, and also ensure that the convergence criteria are not excessively stringent in cases when there is little change in the solution. This talk presents representative multiphysics problems to highlight these issues, and shows strategies for convergence criteria in MOOSE that are robust for multiphysics models under a variety of conditions.

97 - MATHEMATICS AND COMPUTING↗
DOE OSTI2025

Rapid Inference of Logic Gate Neural Networks for Anomaly Detection in High Energy Physics

The increasing data rates and complexity of detectors at the Large Hadron Collider (LHC) necessitate fast and efficient machine learning models, particularly for rapid selection of what data to store, known as triggering. Building on recent work in differentiable logic gates, we present a public implementation of a Convolutional Differentiable Logic Gate Neural Network (CLGN). We apply this to detecting anomalies at the Level-1 Trigger at CMS using public data from the CICADA project. We demonstrate that the CLGN achieves physics performance on par with or superior to conventional quantized neural networks. We also synthesize an LGN for a Field-Programmable Gate Array (FPGA) and show highly promising FPGA characteristics, notably zero Digital Signal Processor (DSP) resource usage. This work highlights the potential of logic gate networks for high-speed, on-detector inference in High Energy Physics and beyond.

FOS: Physical sciences↗