Search NASA⌕ Search

SEARCH · Search NASA

Results for “fine tuning”

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 37 records · Page 2
All sourcesNASA NTRSDOE OSTI
DOE OSTI2026

Data‐Efficient Generation of Synthetic Microstructures of Polymer‐Bonded Energetic Material With Fine‐Tuned Stable Diffusion

Among current deep learning approaches for synthetic image generation, diffusion-based models stand out in terms of algorithmic stability and ability to retain high-fidelity image features with detailed resolution. Here, in this work, we employ Dreambooth, a method for fine-tuning Stable Diffusion, on X-ray CT images of microstructure of the polymer-bonded form (PBX) of a commonly used high explosive, Pentaerythritol tetranitrate (PETN), which yields generative models for creating synthetic PBX images. The models developed here represent five classes (or ‘lots’) of microstructures and demonstrate successful generation of images of each class with high fidelity, as verified by computed classification accuracy of ∼ 94% or higher. Data augmentation afforded by such image synthesis can be used to more reliably decipher underlying statistics, build processing-structure correlations, recognize off-normal structural anomalies, and identify age-related changes. Ideas related to converting image data into appropriate density mapping and performing mesoscale simulation or surrogate modeling of detonation are also discussed.

Dreambooth↗
DOE OSTI2025

Fine‐Tuning Li‐Ion Solvation Structure by Enhanced Solvent‐Diluent Interactions for Long‐Cycling Lithium Metal Batteries

Achieving durable lithium (Li) metal anodes in liquid electrolytes remains challenging, primarily due to the instability of the formed solid-electrolyte interphases (SEIs). Modulating the Li-ion solvation structures is pivotal in forming a stable SEI for stabilizing Li metal anodes. Here a strategy is developed to fine-tune the Li-ion solvation structures through enhanced dipole–dipole interactions between the Li-ion-coordinated solvent and the non-Li-ion-coordinating diluent, for creating a stable SEI in the developed binary salt electrolyte. The enhanced dipole–dipole interactions weaken the coordination between Li-ions and the solvents while strengthening the interaction between Li-ions and dual anions, thereby facilitating the Li-ion transport and a robust anion-derived SEI with a distinct bilayer structure. Consequently, the developed electrolyte exhibited exceptional electrochemical performance in high energy-density Li||LiNi 0.8 Mn 0.1 Co 0.1 O 2 (NMC811) cells, with long calendar life, stable cyclability at 1 C, and reliable operation between 25 and −20 °C, and it also demonstrat remarkable cycling stability for a Li||NMC811 pouch cell with projected energy density of 402 Wh kg −1 , maintaining 80% capacity retention over 606 cycles under practical conditions.

25 ENERGY STORAGE↗
DOE OSTI2024

Fine-tuning TrailMap: The utility of transfer learning to improve the performance of deep learning in axon segmentation of light-sheet microscopy images

Light-sheet microscopy has made possible the 3D imaging of both fixed and live biological tissue, with samples as large as the entire mouse brain. However, segmentation and quantification of that data remains a time-consuming manual undertaking. Machine learning methods promise the possibility of automating this process. This study seeks to advance the performance of prior models through optimizing transfer learning. We fine-tuned the existing TrailMap model using expert-labeled data from noradrenergic axonal structures in the mouse brain. By changing the cross-entropy weights and using augmentation, we demonstrate a generally improved adjusted F1-score over using the originally trained TrailMap model within our test datasets.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

Fine-Tuning Microporosity of Crystalline Vanadomolybdate Frameworks for Selective Adsorptive Separation of Kr from Xe

Selective adsorptive capture and separation of chemically inert krypton (Kr) and xenon (Xe) noble gases with very low ppmv concentrations in air and industrial off-gases constitute an important technological challenge. Here, using a synergistic combination of experiment and theory, the microporous crystalline vanadomolybdates (MoVO x ) as highly selective Kr sorbents are studied in detail. By varying the Mo/V ratios, we show for the first time that their one-dimensional (1D) pores can be fine-tuned for the size-selective adsorption of Kr over the larger Xe with selectivities reaching >100. Using extensive electronic structure calculations and grand canonical Monte Carlo simulations, the competition between Kr uptake with CO 2 and N 2 was also investigated. As most materials reported so far are selective toward the larger, more polarizable Xe than Kr, this work constitutes an important step toward robust Kr-selective sorbent materials. Furthermore this work highlights the potential use of porous crystalline transition metal oxides as energy-efficient and selective noble gas capture sorbents for industrial applications.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗
DOE OSTI2024

Secure Federated Learning Across Heterogeneous Cloud and High-Performance Computing Resources: A Case Study on Federated Fine-Tuning of LLaMA 2

Federated learning enables multiple data owners to collaboratively train robust machine learning models without transferring large or sensitive local datasets by only sharing the parameters of the locally trained models. Here, in this article, we elaborate on the design of our Advanced Privacy-Preserving Federated Learning (APPFL) framework, which streamlines end-to-end secure and reliable federated learning experiments across cloud computing facilities and high-performance computing resources by leveraging Globus Compute, a distributed function as a service platform, and Amazon Web Services. We further demonstrate the use case of APPFL in fine-tuning an LLaMA 2 7B model using several cloud resources and supercomputers.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS1990

Fine-tuning of process conditions to improve product uniformity of polystyrene particles used for wind tunnel velocimetry

Monodisperse polymer particles (having uniform diameter) were used for the last two decades in physical, biological, and chemical sciences. In NASA Langley Research Center monodisperse polystyrene particles are used in wind tunnel laser velocimeters. These polystyrene (PS) particles in latex form were formulated at the Engineering Laboratory of FENGD using emulsion-free emulsion polymerization. Monodisperse PS latices particles having different particle diameters were formulated and useful experimental data involving effects of process conditions on particle size were accumulated. However, similar process conditions and chemical recipes for polymerization of styrene monomer have often yielded monodisperse particles having varying diameters. The purpose was to improve the PS latex product uniformity by fine-tuning the process parameters based on the knowledge of suspension and emulsion polymerization.

Ray, Asit K.↗
DOE OSTI2025

Atomically Fine-Tuning Organic–Inorganic Carbon Molecular Sieve Membranes for Hydrogen Production

Polymeric membranes with great processability are attractive for the H 2 /CO 2 separation required for hydrogen production from renewable biomass with carbon capture for utilization and sequestration. However, it remains elusive to engineer polymer architectures to obtain desired sub-3.3 Å ultramicropores to efficiently sieve H 2 from CO 2 . Herein, we demonstrate a scalable way of carbonizing polybenzimidazole (PBI) at low temperatures, followed by vapor phase infiltration (VPI) to atomically narrow ultramicropores throughout the films, forming hybrid organic–inorganic carbon molecular sieves (CMSs). One VPI cycle (100 s) for the PBI carbonized at 500 °C remarkably increases H 2 /CO 2 selectivity from 9.6 to 83 at 100 °C, surpassing Robeson’s upper bound. The CMS demonstrates a stable H 2 /CO 2 separation performance when challenged with simulated syngas streams and can be fabricated into thin-film composite membranes, outperforming state-of-the-art membranes. Finally, the scalable approach can be ubiquitous to molecularly fine-tune ultramicropores of leading polymeric membranes to further improve their size-sieving ability and thus separation efficiency.

36 MATERIALS SCIENCE↗
DOE OSTI2023

Fine‐Tuning Alkyl Chains on Quinoxaline Nonfullerene Acceptors Enables High‐Efficiency Ternary Organic Solar Cells with Optimizing Molecular Stacking and Reducing Energy Loss

Abstract Material design of guest acceptor is always a big challenge for improving the efficiency of ternary organic solar cells (OSCs). Here, a pair of isomeric nonfullerene acceptors based on quinoxaline core, Qx–p‐C 7 H 8 O and Qx–m‐C 7 H 8 O, is designed and synthesized. By moving the alkoxy chain attached on side phenyl frommeta‐position topara‐position, both π–π stacking distance and crystallinity are enhanced simultaneously. They obtain the uplifted lowest unoccupied molecular orbital level. Compared to Qx–m‐C 7 H 8 O, Qx–p‐C 7 H 8 O exhibits wider absorption spectrum and higher extinction coefficient. Using D18‐Cl:N3 as host materials, the addition of guest acceptor Qx–p‐C 7 H 8 O significantly improves the power conversion efficiency (PCE) from 17.61% to 18.49% because of higher open‐circuit voltage (0.875 V) and short‐circuit current density (27.85 mA cm −2 ). This can be attributed to the faster exciton dissociation, more balanced carrier mobility, fine fiber morphology, and lower energy loss in the ternary devices. However, Qx–m‐C 7 H 8 O‐based ternary device achieves relatively low PCE of 17.17% because this device shows extremely low electron mobility. The results indicate that molecular stacking, film morphology, etc., can be effectively modulated by fine‐tuning the side chains of guest materials, which may be an effective design rule for further improving the PCE of OSCs.

Chemistry↗
DOE OSTI2022

Fine-tuning of mTOR signaling by the UBE4B-KLHL22 E3 ubiquitin ligase cascade in brain development

ABSTRACT Spatiotemporal regulation of the mechanistic target of rapamycin (mTOR) pathway is pivotal for establishment of brain architecture. Dysregulation of mTOR signaling is associated with a variety of neurodevelopmental disorders. Here, we demonstrate that the UBE4B-KLHL22 E3 ubiquitin ligase cascade regulates mTOR activity in neurodevelopment. In a mouse model with UBE4B conditionally deleted in the nervous system, animals display severe growth defects, spontaneous seizures and premature death. Loss of UBE4B in the brains of mutant mice results in depletion of neural precursor cells and impairment of neurogenesis. Mechanistically, UBE4B polyubiquitylates and degrades KLHL22, an E3 ligase previously shown to degrade the GATOR1 component DEPDC5. Deletion of UBE4B causes upregulation of KLHL22 and hyperactivation of mTOR, leading to defective proliferation and differentiation of neural precursor cells. Suppression of KLHL22 expression reverses the elevated activity of mTOR caused by acute local deletion of UBE4B. Prenatal treatment with the mTOR inhibitor rapamycin rescues neurogenesis defects in Ube4b mutant mice. Taken together, these findings demonstrate that UBE4B and KLHL22 are essential for maintenance and differentiation of the precursor pool through fine-tuning of mTOR activity.

Kong, Xiangxing↗
DOE OSTI2020

Chloroplast lipid biosynthesis is fine-tuned to thylakoid membrane remodeling during light acclimation

Abstract Reprogramming metabolism, in addition to modifying the structure and function of the photosynthetic machinery, is crucial for plant acclimation to changing light conditions. One of the key acclimatory responses involves reorganization of the photosynthetic membrane system including changes in thylakoid stacking. Glycerolipids are the main structural component of thylakoids and their synthesis involves two main pathways localized in the plastid and the endoplasmic reticulum (ER); however, the role of lipid metabolism in light acclimation remains poorly understood. We found that fatty acid synthesis, membrane lipid content, the plastid lipid biosynthetic pathway activity, and the degree of thylakoid stacking were significantly higher in plants grown under low light compared with plants grown under normal light. Plants grown under high light, on the other hand, showed a lower rate of fatty acid synthesis, a higher fatty acid flux through the ER pathway, higher triacylglycerol content, and thylakoid membrane unstacking. We additionally demonstrated that changes in rates of fatty acid synthesis under different growth light conditions are due to post-translational regulation of the plastidic acetyl-CoA carboxylase activity. Furthermore, Arabidopsis mutants defective in one of the two glycerolipid biosynthetic pathways displayed altered growth patterns and a severely reduced ability to remodel thylakoid architecture, particularly under high light. Overall, this study reveals how plants fine-tune fatty acid and glycerolipid biosynthesis to cellular metabolic needs in response to long-term changes in light conditions, highlighting the importance of lipid metabolism in light acclimation.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2022

FINETUNA: fine-tuning accelerated molecular simulations

Abstract Progress towards the energy breakthroughs needed to combat climate change can be significantly accelerated through the efficient simulation of atomistic systems. However, simulation techniques based on first principles, such as density functional theory (DFT), are limited in their practical use due to their high computational expense. Machine learning approaches have the potential to approximate DFT in a computationally efficient manner, which could dramatically increase the impact of computational simulations on real-world problems. However, they are limited by their accuracy and the cost of generating labeled data. Here, we present an online active learning framework for accelerating the simulation of atomic systems efficiently and accurately by incorporating prior physical information learned by large-scale pre-trained graph neural network models from the Open Catalyst Project. Accelerating these simulations enables useful data to be generated more cheaply, allowing better models to be trained and more atomistic systems to be screened. We also present a method of comparing local optimization techniques on the basis of both their speed and accuracy. Experiments on 30 benchmark adsorbate-catalyst systems show that our method of transfer learning to incorporate prior information from pre-trained models accelerates simulations by reducing the number of DFT calculations by 91%, while meeting an accuracy threshold of 0.02 eV 93% of the time. Finally, we demonstrate a technique for leveraging the interactive functionality built in to Vienna ab initio Simulation Package (VASP) to efficiently compute single point calculations within our online active learning framework without the significant startup costs. This allows VASP to work in tandem with our framework while requiring 75% fewer self-consistent cycles than conventional single point calculations. The online active learning implementation, and examples using the VASP interactive code, are available in the open source FINETUNA package on Github.

97 MATHEMATICS AND COMPUTING↗
NASA NTRSDate not supplied

Towards an Aviation Large Language Model by Fine-tuning and Evaluating Transformers

In the aviation domain, there are many applications for machine learning and artificial intelligence tools that utilize natural language. For example, there is a desire to know the commonalities in written safety reports such as voluntary post incidents reports or create more accurate transcripts of air traffic management conversations. Another use-case is the possibility of extracting airspace procedures and constraints currently written in documents such as Letters of Agreement (LOA) which is used as the evaluation case in this paper. These applications can benefit from the use of state-of-the-art Natural Language Processing (NLP) techniques when adapted to the language/phraseology specific to the aviation domain. This paper evaluates the viability of transferring pre-trained large language models to the aviation domain by adapting transformer based models using aviation datasets. This paper utilized two datasets to adapt a ‘Robustly Optimized Bidirectional Encoder Representations from Transformers Approach’ (RoBERTa) model and two down-stream classification tasks to assess its performance. These datasets are all built upon Letters of Agreement which are Federal Aviation Administration (FAA) documents that formalize airspace operations across the national airspace system. The first two datasets are used for the adaptation of RoBERTa to the aviation domain and were of different sizes to assess the number of documents needed to adapt to the aviation domain. They contain many examples of ‘aviation English’ using domain specific terminology and phrasing which serves as a representative basis to perform the unsupervised adaptation. The second dataset is a separate set of LOA documents with two sets of classification labels to be used for evaluation; one at the document level and one at the line level. These down-stream evaluations allowed the measurement of improvement by adapting RoBERTa. The accuracy increased by 4-6% on both tasks and the F1 score on the class of interest increased by 4-8% from the adaptation.

Air Traffic Management↗
NASA NTRSDate not supplied

Towards an Aviation Large Language Model by Fine-tuning and Evaluating Transformers

In the aviation domain, there are many applications for machine learning and artificial intelligence tools that utilize natural language. For example, there is a desire to know the commonalities in written safety reports such as voluntary post incidents reports or create more accurate transcripts of air traffic management conversations. Another use-case is the possibility of extracting airspace procedures and constraints currently written in documents such as Letters of Agreement (LOA) which is used as the evaluation case in this paper. These applications can benefit from the use of state-of-the-art Natural Language Processing (NLP) techniques when adapted to the language/phraseology specific to the aviation domain. This paper evaluates the viability of transferring pre-trained large language models to the aviation domain by adapting transformer based models using aviation datasets. This paper utilized two datasets to adapt a ‘Robustly Optimized Bidirectional Encoder Representations from Transformers Approach’ (RoBERTa) model and two down-stream classification tasks to assess its performance. These datasets are all built upon Letters of Agreement which are Federal Aviation Administration (FAA) documents that formalize airspace operations across the national airspace system. The first two datasets are used for the adaptation of RoBERTa to the aviation domain and were of different sizes to assess the number of documents needed to adapt to the aviation domain. They contain many examples of ‘aviation English’ using domain specific terminology and phrasing which serves as a representative basis to perform the unsupervised adaptation. The second dataset is a separate set of LOA documents with two sets of classification labels to be used for evaluation; one at the document level and one at the line level. These down-stream evaluations allowed the measurement of improvement by adapting RoBERTa. The accuracy increased by 4-6% on both tasks and the F1 score on the class of interest increased by 4-8% from the adaptation.

Air Traffic Management↗
NASA NTRS1985

Resistive Fine Tuning of Resonant Circuit

Simple fixed-inductance/fixed-capacitance tank circuit modified for fine adjustment of resonant frequency by addition of small inductance with potentiometer across it. Additional winding built into full winding as integral part or added externally. Technique provides quick way of tuning reactance out of power-transformer circuit to maximize power transfer or to adjust frequency of oscillator. Applications include rotary transformers, servo amplifiers, and analog computer modules.

Mclyman, C. W.↗
DOE OSTI2023

19.32% Efficiency Polymer Solar Cells Enabled by Fine-Tuning Stacking Modes of Y-Type Molecule Acceptors: Synergistic Bromine and Fluorine Substitution of the End Groups

Here, the success of Y6-type nonfullerene small molecule acceptors (NF-SMAs) in polymer solar cells (PSCs) can be attributed to their unique honeycomb stacking style, which leads to favorable thin-film morphologies. The intermolecular interactions related to the crystallization tendency of these NF-SMAs is closely governed by their electron accepting end groups. For example, the high performance Y6 derivative L8-BO (BTP-4F) presents three types of stacking modes in contrast to two stacking modes of Y6. Hence, it is ultimately interesting to obtain more insight on the packing properties and the preferences influenced by chemical modifications such as end group engineering. This work designs and synthesizes asymmetric and symmetric L8-BO derivatives with brominated end groups and explores the stacking preferences in various modes. The asymmetric BTP-3FBr displays an optimized crystallization tendency and thin film morphology, leading to a decent power conversion efficiency (PCE) of 18.34% in binary devices and a top PCE of 19.32% in ternary devices containing 15 wt% IDIC as the second acceptor.

36 MATERIALS SCIENCE↗
DOE OSTI2026

Theory Guided Fine‐Tune of Strain Effects in Pt Ternary Alloy via Rare Earth Templating: Achieving High Performance PEMFCs Catalysts

The sluggish kinetics and insufficient durability of platinum-based catalysts remain crucial barriers limiting proton-exchange-membrane fuel cells (PEMFCs) deployment. Here, we report a theory-guided synthesis combined with rare-earth templating to realize a previously inaccessible Pt 5 Co-like phase with tailored atomic-scale strain. Guided by density functional theory (DFT) calculations, we identified that a Pt 5 Co-like sublayer can induce a unique mild compressive strain (−1.24%) to the Pt(111) shell and an optimal *OH binding energy shift (ΔE ≈ 0.11 eV). This shift positions the alloy catalyst near the apex of the oxygen reduction reaction activity volcano. This prediction guided the synthesis of ternary alloy Pt 5 (Ce)Co@Pt multilayer nanoparticles, featuring a Ce-stabilized core, a Pt 5 Co-like sublayer, and a Pt-rich shell. This catalyst demonstrates both exceptionally high activity and durability, achieving a mass activity of 2.6 A∙mg Pt −1 in rotating disk electrode testing. In fuel cell membrane electrode assembly tests, Pt 5 (Ce)Co@Pt achieves a current density of 1.9 A∙cm −2 at 0.7 V under heavy-duty vehicle conditions. Remarkably, it maintains 1.2 A∙cm −2 after 1 80 000 AST cycles, doubling the U.S. DOE 2025 target. This work demonstrates a rational design strategy that DFT-guided strain engineering integrates with rare-earth templating to advance Pt-based catalysts for fuel cell applications.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗
DOE OSTI2022

Fine-tuned vs. natural supersymmetry: what does the string landscape predict?

A vast array of (metastable) vacuum solutions arise from string compactifications, each leading to different 4-d laws of physics. The space of these solutions, known as the string landscape, allows for an environmental solution to the cosmological constant problem. We examine the possibility of an environmental solution to the gauge hierarchy problem. We argue that the landscape favors softly broken supersymmetric models over particle physics models containing quadratic divergences, such as the Standard Model. We present a scheme for computing relative probabilities for supersymmetric models to emerge from the landscape. The probabilities are related to the likelihood that the derived value of the weak scale lies within the Agrawal et al. (ABDS) allowed window of values leading to atoms as we know them. This then favors natural SUSY models over unnatural (SUSY and other) models via a computable probability measure.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗