Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “predictability”

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 1,729 records · Page 96
All sourcesNASA NTRSDOE OSTI
DOE OSTI2025

Prediction of an alternative high-pressure route to polymeric carbon dioxide as a metastable energetic material

The use of pressure to obtain new materials that can be recovered under ambient conditions is a central problem in high-pressure physics. Despite decades of research, this goal has only been achieved in the laboratory for a few notable examples, such as diamond and cubic boron nitride. An area of significant interest is the transformation under compression of light-element molecular compounds to extended covalent-bonded (polymeric) solids. Among them, CO 2 has been extensively studied because of its status as a prototypical simple molecular system with a rich phase diagram and due to its fundamental role in Earth’s physics and chemistry. One of its polymeric crystalline phases, accessible at extreme pressures and temperatures, has been recently quenched to ambient pressure, but below room temperature. Here we report ab initio calculations predicting that isothermal compression of a carbon monoxide and oxygen mixture (CO+O 2 ), rather than the compound CO 2 , lowers the onset of C-polymerization at room temperature from ~ 118 GPa to ~ 7 GPa (complete by ~ 23 GPa). Moreover, it leads to the formation of an intrinsically different polymer with enhanced metastability. We predict that this dense phase is an energetic material which can potentially be recovered to ambient pressure and temperature.

Paul, Reetam [Lawrence Livermore National Laborato↗
DOE OSTI2021

Predicting 1 H NMR relaxation in Gd 3+ -aqua using molecular dynamics simulations

Atomistic molecular dynamics simulations are used to predict 1 H NMR T 1 relaxation of water from paramagnetic Gd 3+ ions in solution at 25 °C. Simulations of the T 1 relaxivity dispersion function r 1 computed from the Gd 3+ – 1 H dipole–dipole autocorrelation function agree within ≃8% of measurements in the range f 0 ≃ 5 ↔ 500 MHz, without any adjustable parameters in the interpretation of the simulations, and without any relaxation models. Here the simulation results are discussed in the context of the Solomon-Bloembergen-Morgan inner-sphere relaxation model, and the Hwang-Freed outer-sphere relaxation model. Below f 0 ≲ 5 MHz, the simulation overestimates r 1 compared to measurements, which is used to estimate the zero-field electron-spin relaxation time. The simulations show potential for predicting r 1 at high frequencies in chelated Gd 3+ contrast-agents used for clinical MRI.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2022

The solution structures and relative stability constants of lanthanide–EDTA complexes predicted from computation

Ligand selectivity to specific lanthanide (Ln) ions is key to the separation of rare earth elements from each other. Ligand selectivity can be quantified with relative stability constants (measured experimentally) or relative binding energies (calculated computationally). The relative stability constants of EDTA (ethylenediaminetetraacetic acid) with La 3+ , Eu 3+ , Gd 3+ , and Lu 3+ were predicted from relative binding energies, which were quantified using electronic structure calculations with relativistic effects and based on the molecular structures of Ln–EDTA complexes in solution from density functional theory molecular dynamics simulations. The protonation state of an EDTA amine group was varied to study pH ~7 and ~11 conditions. Further, simulations at 25 °C and 90 °C were performed to elucidate how structures of Ln–EDTA complexes varying with temperature are related to complex stabilities at different pH conditions. Relative stability trends are predicted from computation for varying Ln 3+ ions (La, Eu, Gd, Lu) with a single ligand (EDTA at pH ~11), as well as for a single Ln 3+ ion (La) with varying ligands (EDTA at pH ~7 and ~11). Changing the protonation state of an EDTA amine site significantly changes the solution structure of the Ln–EDTA complex resulting in a reduction of the complex stability. As a result, increased Ln–ligand complex stability is correlated to reduced structural variations in solution upon an increase in temperature.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2022

Predicting third-body collision efficiencies for water and other polyatomic baths

Low-pressure-limit microcanonical (collisional activation) and thermal rate constants are predicted using a combination of automated ab initio potential energy surface construction, classical trajectories, transition state theory, and a detailed energy- and angular-momentum-resolved collision kernel. Several systems are considered, including CH 4 (+M) and HO 2 (+M), with an emphasis on systems where experimental information is available for comparison. The a priori approach involves no adjustable parameters, and we show that the predicted thermal rate constants are in excellent agreement with experiments, with average deviations of less than 25%. Notably, the a priori approach is shown to perform equally well for atomic, diatomic, and polyatomic baths, including M = H 2 O, CO 2 , and “fuel” baths like M = CH 4 and NH 3 . Additionally, the utility of microcanonical rate constants for interpreting trends and inferring mechanistic details in the thermal kinetics is demonstrated.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2023

Predicting and accessing metastable phases

Metastable phase formation of Ln 2 O 3 phases was explained and successfully predicted using calculated metastable phase diagrams. Irradiation experiments of Lu 2 O 3 show formation of 3 phases, confirming the prediction, showing a unique behavior.

36 MATERIALS SCIENCE↗
DOE OSTI2023

Large-scale computational polymer solubility predictions and applications to dissolution-based plastic recycling

Dissolution-based plastic recycling is a promising approach to separate and recover high quality pure polymer resins from multicomponent plastic waste by exploiting differences in polymer solubility. The design of a dissolution-based polymer recycling process requires the selection of appropriate solvent systems and operating temperatures to dissolve only target polymers. Determining these parameters experimentally is challenging due to the wide range of solvents and temperatures possible for a given set of target polymers. In this work, we report a computational scheme that employs molecular dynamics simulations and the Conductor-like Screening Model for Realistic Solvents to predict polymer solubilities. Using this scheme, we established a computational solubility database for 8 common polymers and 1007 solvents at multiple temperatures and measured selected solubilities experimentally to validate computational predictions. Analysis of functional groups within this large database then provides chemical heuristics relating the molecular structures of good and non-solvents for selected polymers. We further developed a tool that automates the selection of solvents for all possible sequences in which target polymers can be selectively dissolved to guide the design of dissolution-based plastic recycling processes. Here, we demonstrate the application of these methods via multiple experimental case studies of representative dissolution-based polymer recycling processes in which pure polymer resins were successfully recovered from physical mixtures of polymers.

42 ENGINEERING↗
DOE OSTI2023

Machine learning predictions of diffusion in bulk and confined ionic liquids using simple descriptors

Ionic liquids have many intriguing properties and widespread applications such as separations and energy storage. However, ionic liquids are complex fluids and predicting their behavior is difficult, particularly in confined environments. We introduce fast and computationally efficient machine learning (ML) models that can predict diffusion coefficients and ionic conductivity of bulk and nanoconfined ionic liquids over a wide temperature range (350–500 K). The ML models are trained on molecular dynamics simulation data for 29 unique ionic liquids as bulk fluids and confined in graphite slit pores. This model is based on simple physical descriptors of the cations and anions such as molecular weight and surface area. Here, we also demonstrate that accurate results can be obtained using only descriptors derived from SMILES (simplified molecular-input line-entry system) codes for the ions with minimal computational effort. This offers a fast and efficient method for estimating diffusion and conductivity of nanoconfined ionic liquids at various temperatures without the need for expensive molecular dynamics simulations.

74 ATOMIC AND MOLECULAR PHYSICS↗
DOE OSTI2025

SLAB: simultaneous labeling and binding affinity prediction for protein–ligand structures

Machine learning models are often used as scoring functions to predict the binding affinity of a protein–ligand complex. These models are trained with limited amounts of data with experimentally measured binding affinity values. A large number of compounds are labeled inactive through single-concentration screens without measuring binding affinities. These inactive compounds, along with the active ones, can be used to train binary classification models, while regression models are trained using compounds with binding affinities only. However, the classification and regression tasks are often handled separately, without sharing the learned feature representations. In this paper, we propose a novel model architecture that jointly performs regression and classification objectives, aiming to maximize data utilization and improve predictive performance by leveraging two complementary tasks. In our setup, the regression yields the binding affinity, whereas the classification task yields the label as active or inactive. We demonstrate our method using PDBbind, the standard 3D structure database, as well as a dataset of flavivirus protease compounds with binding affinity data. Our experiments show that the new joint training strategy improves the accuracy of the model, increasing applicability in various practical drug screening scenarios.

Biological and medical sciences↗
DOE OSTI2025

Molecular property prediction for very large databases with natural language processing: a case study in ionic liquid design

The prospect of using artificial intelligence (AI) to accurately screen very large databases of compounds for multiple properties has yet to be realized. Here, we explore this possibility using ionic liquids (ILs) which offer unique physicochemical properties and excellent tunability, making them highly versatile solvents for various research applications. Screening millions of potential ILs for the best perfomance for use in specific tasks with experimental methods alone however, is impractical. Further, traditional’ physics-based computational chemistry is hindered by high computational cost. To address this challenge, we leverage a natural language processing (NLP)-based molecular embedding technique with advanced machine learning (ML) models to predict seven key IL properties: viscosity, density, ionic conductivity, surface tension, melting temperature, toxicity, and water solubility. Comprehensive datasets for these properties are obtained, then NLP featurization with Mol2vec is compared with other featurization techniques such as 2D Morgan fingerprints, and 3D quantum chemistry-derived sigma profiles. NLP-based featurization exhibited the best predictive performance, achieving the highest R 2 and lowest RMSE values for all the studied IL properties. Further, we present case studies of how ILs might be screened using combined property criteria for practical cases – lignocellulosic biomass processing, CO 2 capture, and optimal electrolytes for batteries – screening a novel database of ∼10.6 million generated feasible ILs. The results introduce NLP as a powerful tool for engineering many designer solvents with desirable properties for task specific applications.

Mohan, Mood [Oak Ridge National Laboratory (ORNL),↗
DOE OSTI2026

Structure prediction of porous organic crystals

In this work, we explore the possibility of applying automated crystal structure prediction to reproduce the experimentally identified metastable porous polymorphs. Using our recently developed High-Throughput Organic Crystal Structure Prediction ( HTOCSP ) framework, we conducted a systematic study on five representative organic crystalline systems including hydrogen-bonded frameworks (HOFs), featured by the presence of significant porosity, in conjunction with different choices of energy models from classical, machine learning force fields, tight binding to density functional theory. Our results suggest that the current structure generation framework, with careful selection of symmetry conditions, is likely to generate rather complex and abundant metastable crystal candidates for porous crystals. In conjunction with the recent advance in universal machine learning force fields, it becomes possible to identify experimental structures as the energetically favorable candidates from a simple energy versus density analysis, thus paving the way for computational design of complex porous materials with the target systems prior to the experimental synthesis and characterization.

36 MATERIALS SCIENCE↗
DOE OSTI2026

Integrated machine learning-molecular dynamics framework for electrolyte property prediction

Electrochemical stability windows determine the operating range of battery electrolytes, yet accurate prediction remains challenging because stability emerges from statistical ensembles of local solvation environments rather than single ground-state molecular structures. Traditional density functional theory calculations on energy-minimized clusters cannot capture the thermal variations in local coordination environments and geometries that govern decomposition, while SMILES-based machine learning methods lack explicit representation of three-dimensional solvation structure and ion pairing. Here, we introduce a structure-aware machine learning framework that predicts frontier orbital energies (HOMO and LUMO) directly from molecular dynamics-sampled solvation configurations, achieving sub-0.6 eV accuracy at computational costs 3–4 orders of magnitude lower than first-principles methods. Across twelve representative battery electrolytes, we demonstrate that solvent-separated and contact ion pairs exhibit strong size- and local chemistry dependent electronic stability, with variations in coordination shifts of HOMO or LUMO level by 2–3 eV, and that extended solvation structure and partially desolvated environment further modulate stability by up to 3 eV. By encoding the statistical nature of electrochemical failure through ensemble sampling of explicit solvation geometries, our approach enables high-throughput screening and rational design of next-generation battery electrolytes with mechanistic understanding of structure–property relationships.

Energy - Storage↗
DOE OSTI2025

ForestFlow: predicting the Lyman-α forest clustering from linear to nonlinear scales

On large scales, the Lyman-α forest provides insights into the expansion history of the Universe, while on small scales, it imposes strict constraints on the growth history, the nature of dark matter, and the sum of neutrino masses. This work introduces ForestFlow, a novel framework that bridges the gap between large- and small-scale analyses, which have traditionally relied on distinct modeling approaches. Using conditional normalizing flows, ForestFlow predicts the two Lyman-α linear biases (b δ and b η ) and six parameters describing small-scale deviations of the three-dimensional flux power spectrum (P 3D ) from linear theory as a function of cosmology and intergalactic medium physics. These are then combined with a Boltzmann solver to make consistent predictions, from arbitrarily large scales down to the nonlinear regime, for P 3D and any other statistics derived from it. Trained on a suite of 30 fixed-and-paired cosmological hydrodynamical simulations spanning redshifts from z = 2 to 4.5, ForestFlow achieves 3 and 1.5% precision in describing P 3D and the one-dimensional flux power spectrum (P 1D ) from linear scales to k = 5 Mpc −1 and k ∥ = 4 Mpc −1 , respectively. Thanks to its conditional parameterization, ForestFlow shows similar performance for ionization histories and two ΛCDM model extensions – massive neutrinos and curvature – even though none of these are included in the training set. This framework will enable full-scale cosmological analyses of Lyman-α forest measurements from the DESI survey.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2024

Predicting nucleon-nucleus scattering observables using nuclear structure theory

Developing a predictive capability for inelastic scattering will find applications in multiple areas. Experimental data for neutron-nucleus inelastic scattering is limited and thus one needs a robust theoretical framework to complement it. Charged-particle inelastic scattering can be used as a surrogate for ( n, γ ) reactions to predict capture cross sections for unstable nuclei. Our work uses microscopic nuclear structure calculations for spherical nuclei to obtain nucleon-nucleus scattering potentials and calculate cross sections for these processes. We implement the Jeukenne, Lejeune, Mahaux (JLM) semi-microscopic folding approach, where the medium effects on nuclear interaction are parameterized in nuclear matter to obtain the nucleon-nucleon ( NN ) interaction in a medium at positive energies. We solve for the nuclear ground state using the Hartree-Fock-Bogliubov (HFB) many-body method, assuming the nucleons within the nucleus interact via the Gogny-D1M potential. The vibrational excited states of the target nucleus are calculated using the quasi-particle random phase approximation (QRPA). We demonstrate our approach for spherical nuclei in the medium-mass region, showing scattering results for the 90 Zr nucleus.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
DOE OSTI2020

Predicting metal–metal interactions. I. The influence of strain on nanoparticle and metal adlayer stabilities

Strain-engineering of bimetallic nanomaterials is an important design strategy for developing new catalysts. In this paper, we introduce an approach for including strain effects into a recently introduced, density functional theory (DFT)-based alloy stability model. The model predicts adsorption site stabilities in nanoparticles and connects these site stabilities with catalytic reactivity and selectivity. Strain-based dependencies will increase the model’s accuracy for nanoparticles affected by finite-size effects. In addition to the stability of small nanoparticles, strain also influences the heat of adsorption of epitaxially grown metal-on-metal adlayers. In this respect, we successfully benchmark the strain-including alloy stability model with previous experimentally determined trends in the heats of adsorption of Au and Cu adlayers on Pt (111). For these systems, our model predicts stronger bimetallic interactions in the first monolayer than monometallic interactions in the second monolayer. We explicitly quantify the interplay between destabilizing strain effects and the energy gained by forming new metal–metal bonds. While tensile strain in the first Cu monolayer significantly destabilizes the adsorption strength, compressive strain in the first Au monolayer has a minimal impact on the heat of adsorption. Hence, this study introduces and, by comparison with previous experiments, validates an efficient DFT-based approach for strain-engineering the stability, and, in turn, the catalytic performance, of active sites in bimetallic alloys with atomic level resolution.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2020

Deep convolutional neural networks for multi-scale time-series classification and application to tokamak disruption prediction using raw, high temporal resolution diagnostic data

In this paper we discuss recent advances in deep convolutional neural networks (CNN) for sequence learning, which allow identifying long-range, multi-scale phenomena in long sequences, such as those found in fusion plasmas. We point out several benefits of these deep CNN architectures, such as not requiring experts such as physicists to hand-craft input data features, the ability to capture longer range dependencies compared to the more common sequence neural networks (recurrent neural networks like long short-term memory (LSTM) networks), and the comparative computational efficiency. We apply this neural network architecture to the popular problem of disruption prediction in fusion energy tokamaks, utilizing raw data from a single diagnostic, the Electron Cyclotron Emission imaging (ECEi) diagnostic from the DIII-D tokamak. Initial results trained on a large ECEi dataset show promise, achieving an F 1 -score of ~91% on individual time-slices using only the ECEi data. This indicates the ECEi diagnostic by itself can be sensitive to a number of pre-disruption markers useful for predicting disruptions on timescales not only for mitigation but also avoidance. Future opportunities for utilizing these deep CNN architectures with fusion data are outlined, including impact of recent upgrades to the ECEi diagnostic.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2023

Time-temperature-stress superposition of PBX 9502 compressive creep data for lifetime predictions

The plastic-bonded explosive (PBX) 9502 is a viscoelastic, high-solids loaded polymer bound composite comprised of 95 weight% (wt%) TATB explosive crystals and 5 wt% FK-800 polymer binder. The uniaxial quasi-static mechanical properties have been studied and characterized extensively over 25+ years, including creep. Creep is the strain evolution over time as a result of an asymmetric load typically at temperatures above ambient. Creep measurements to failure can take excessively long periods of time depending on various factors including applied stress, temperature, and loading rate that render a single test to failure impossible. Using the principles of time-temperature-stress superposition (TTSSP), short-term creep tests at various applied stresses and temperatures can be used for long-term creep predictions. The TTSSP model shifts short-term test data using two factors, one for temperature and stress, creating a master curve at a reference test condition. The master curve can be shifted to predict the creep response at any desired test condition. PBX 9502 compression specimens were tested at three temperatures and various applied stresses. This work discusses the TTSSP analysis of PBX 9502 compressive creep data to create a single comprehensive master curve.

36 MATERIALS SCIENCE↗