Search NASA⌕ Search

SEARCH · Search NASA

Results for “Partial Optimization”

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 199 records · Page 11
All sourcesNASA NTRSDOE OSTI
DOE OSTI2021

Exploring the Meta-regulon of the CRP/FNR Family of Global Transcriptional Regulators in a Partial-Nitritation Anammox Microbiome

Microbiomes are important contributors to many ecosystems, including ones where nutrient cycling is stimulated by aeration control. Optimizing cyclic aeration helps reduce energy needs and maximize microbiome performance during wastewater treatment; however, little is known about how most microbial community members respond to these alternating conditions.

transcription factors↗
DOE OSTI2025

Optimizing Insulation Design for Transformers in Medium Voltage Power Conversion Systems

Medium-frequency transformers (MFTs) play a crucial role in medium-voltage (MV) solidstate transformer (SST) systems, particularly in extreme fast charging applications. Achieving partial discharge (PD)-free operation while maintaining high power density is a significant challenge due to the high electric field (E-field) stresses inherent in MV applications. This dissertation focuses on the insulation design and optimization of MFTs used in both the main power electronics circuits and auxiliary power supplies. The study begins with an overview of insulation testing methodologies, including high potential tests, basic insulation level tests, and PD tests, which are critical for evaluating MFT insulation reliability. Given the importance of PD-free operation for long-term reliability, particular emphasis is placed on understanding PD mechanisms, including void, corona, and surface discharge, and their mitigation strategies. A high voltage isolated auxiliary power supply is then introduced, utilizing a gapped transformer encapsulated in silicone gel. This design achieves PD-free insulation up to 18 kV RMS while maintaining low coupling capacitance to minimize common-mode current. The proposed solution ensures reliable operation in MV environments and offers a scalable approach for auxiliary power in cascaded SST architectures. To improve MFT insulation in main power conversion circuits, a novel structure is developed using polypropylene sheets and potting compounds to create a void-free air gap, effectively mitigating E-field intensity. A prototype transformer with this insulation structure is built and achieves PD-free operation up to 30 kV RMS. This design is experimentally validated in a resonant converter operating at 46 kW, demonstrating its feasibility for MV SST applications. Further optimization is implemented to enhance MFT performance for dual-active-bridge(DAB) converters by integrating a semiconductive shielding layer within the insulation structure. This shielding layer improves the magnetic coupling coefficient while effectively confining the E-field within high insulation materials, thereby reducing eddy current losses. The optimized MFT achieves PD-free operation at 12.6 kV RMS and is successfully tested in a DAB converter operating at 43 kW, which meets the insulation requirements for a 13.2 kV SST system. This dissertation advances MFT insulation design by introducing and experimentally validating novel approaches that improve high voltage insulation while optimizing magnetic coupling and manufacturability. The proposed insulation structures enable PD-free operation while minimizing insulation material usage and simplifying assembly, making them ideal for high power, high voltage applications.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗
DOE OSTI2024

Impact of mechanical tolerances on partial turn skipping in helical flux compression generator

Maintaining tight mechanical tolerances of the components used in helical flux compression generators is crucial for optimizing device performance. Moving beyond current literature, which provides various approximations for acceptable tolerances, a 3D simulation of the armature stator interaction during generator operation was developed, revealing that existing equations frequently misestimate the tolerance limits. Without restricting the generality of the presented approach, the focus was primarily on the armature’s concentricity and roundness, comparing the acceptable tolerance limits between generators of different sizes. An analytical approach for select geometries was developed, confirming the 3D results. Quantitative results reveal that a steeper Gurney angle and increased winding pitch allow for less restrictive tolerances for eccentricity and ellipticity. Even small deviations from the ideal armature positioning and shape will result in fluctuations of the output current time-derivative, observed experimentally and in simulation. Larger deviations then cause partial turn skipping, associated with loss of magnetic flux. The simulation elucidates the impact of tolerance deviations on generator performance by offering precise tolerance ranges for various generator sizes, thereby facilitating informed design decisions.

42 ENGINEERING↗
DOE OSTI2022

Coupling gas purging with inorganic carbon supply to enhance biohydrogen production with Clostridium thermocellum

Clostridium thermocellum is a desirable biocatalyst for biohydrogen production, with a native ability to simultaneously saccharify cellulose and to metabolize released cellodextrins for hydrogen production. During fermentation with C. thermocellum , partial pressures of two gases - CO 2 and H 2 - are critical drivers of overall reaction kinetics. Biohydrogen production is enhanced by maintaining a low hydrogen partial pressure, while high concentrations of dissolved CO 2 promote microbial biomass synthesis. Our study evaluates the inherent trade-offs between hydrogen stripping and inorganic carbon supply for optimized biohydrogen synthesis. Here, we find that nitrogen sparging at low flow rates increases hydrogen production when compared with an equivalent nitrogen overlay, but that high rates of nitrogen sparging inhibit cell growth and hydrogen production. Decreasing dissolved hydrogen partial pressure via nitrogen sparging also lowers the production of reduced metabolites, including lactate and ethanol. To address potential stripping of inorganic carbon from the production medium, we supplemented CO 2 to the sparging gas and co-optimized for gas flow rate and for the CO 2 fraction of the sparging gas. Total hydrogen production increased from 50 mmol∙L -1 in the base condition, when the bioreactor was sparged with 0.1 LPM of pure nitrogen, to 181.3 mmol∙L -1 when sparged with 1.3 LPM of 33 % CO 2 , demonstrating that biohydrogen production is highly sensitive to both parameters. Fine sensitivity of biohydrogen production to sparging conditions highlights the critical importance of bioreactor design and operation to achieve maximum H 2 removal without compromising inorganic carbon supply to bacterial central metabolism.

09 BIOMASS FUELS↗
DOE OSTI2026

The solubility and speciation of REE phosphate endmembers (CePO 4 and YPO 4 ) in Cl-rich aqueous fluids from 350 to 450 °C and implications for natural systems

The rare earth elements (REE) are important metals used increasingly in advanced technologies. Within the crust, the elements Ce and Y are commonly more abundant compared to other lanthanides and comprise important end-member constituents of REE-bearing minerals. Specifically, Ce is part of the light (L) REE which have larger ionic radii than the heavy (H) REE, which are grouped together with Y. These differences in ionic radius can lead to important physico-chemical trends within the lanthanide group. Despite a recent increase in experimental and thermodynamic data for the REE at high temperature and pressure, there is still a significant lack of these data at supercritical conditions. In this study we conducted batch-type experiments to measure the solubility of REE phosphates (CePO 4 and YPO 4 ) at varying starting pH (1.5–10), and salinity (0.01–1.4 mol/kg NaCl) at 350 and P sat , and from 400 to 450 °C at 700 bar. Results show that the solubility of Ce (33–0.14 ppb) is generally higher than Y (13–0.13 ppb) and that Ce complexes more strongly with both chloride and hydroxyl ligands compared to Y. The solubilities of both REE phosphates are highly pH-dependent and, to a lesser extent, depend on salinity at the studied conditions. The solubility data from this study were implemented into the GEMSFITS program to optimize the thermodynamic properties of Ce and Y hydroxyl and chloride species. The updated standard partial molal Gibbs energies of formation (Δ f G 0 T,P ) are used within the experimental temperature and pressure range to accurately predict the CePO 4 and YPO 4 solubility and Ce and Y speciation behavior. Based on the updated thermodynamic properties we also provide formation constants (log β n Cl,OH ) for Ce and Y hydroxyl and chloride species. Updated thermodynamic properties are applied to model REE-apatite dissolution and REE mobility based on the Pea Ridge iron oxide apatite deposit in Missouri, USA. The apatite dissolution model replicates natural observations including the replacement of monazite and xenotime after apatite and is an example of the utility of the new thermodynamic constants applied to supercritical crustal fluids. Furthermore, the findings of this study advance the predictive capabilities of geochemical models, our understanding of the behavior of individual REE, and permit modeling the overarching fractionation trends between LREE and HREE in supercritical crustal fluids.

58 GEOSCIENCES↗
DOE OSTI2025

Optimizing CO 2 -Loaded Aqueous Amine Solutions for Higher Electrocatalytic CO 2 Reduction Activity

The activity of aqueous-based carbon dioxide reduction (CO 2 R) reactions is often limited by the solubility of CO 2 . The addition of amines can increase the total dissolved carbon in water through the formation of bicarbonate and carbamate species, which has been used to a great effect to capture CO 2 from dilute streams. Here, in this study, we explore the effect of 12 primary and secondary amines of varying Brønsted basicity, steric profile, and hydrogen-bonding capabilities on the aqueous CO 2 R to CO activity of a molecular Ni(cyclam)Cl 2 catalyst with a Hg electrode. Addition of some of the amines results in greater activity and selectivity for CO production compared to equivalent aqueous solutions without added amines. Under optimal conditions (0.4 M 3-amino-propionitrile), there is an over sevenfold increase in partial current density and greater selectivity for CO compared to equivalent conditions with no amine. Interestingly, the increase in activity did not correlate to any single property across the 12 amines. To elucidate the effect of the amine additives on catalysis, we used vapor–liquid equilibrium modeling (VLE), 13 C NMR spectroscopy, and computational analysis to determine the carbon speciation of the solutions. These results indicate that for amines without ethylalcohol functionalities, CO 2 R activity correlates with carbamate concentration, which is in turn governed by amine basicity and steric effects. However, this correlation does not persist for amines with ethylalcohol functionalities, which can form more stable carbamates through intramolecular-hydrogen bonding. These studies demonstrate that amine additives can enhance aqueous CO 2 R activity and selectivity and describe amine properties that lead to these higher performance metrics.

Amines↗
DOE OSTI2022

Thermodynamic guiding principles of high-capacity phase transformation materials for splitting H 2 O and CO 2 by thermochemical looping

Here, thermochemical looping splitting of water and carbon dioxide (CO 2 ) with greenhouse-gas-free (GHG-free) energy has the potential to help address the Gt-scale GHG emissions challenge. Reaction thermodynamics largely contributes to the main bottlenecks of cost reduction for thermochemical looping water/CO 2 splitting cycle. Here, we analyze thermodynamic driving forces in such cycles with two-phase ternary ferrites as model systems. We find that cation configurational entropy chiefly determines the change of partial molar entropy with oxygen stoichiometry. In addition, our phase diagram analysis accurately predicts the optimal Fe ratio for maximal water/CO 2 splitting capacity in thermal reduction and in chemical reduction based cycles, underlining the significance of phase boundary positions. With chemical reduction, >10% CO 2 conversion and high oxygen exchange capacity can both be achieved. Furthermore, our reduced Gibbs free energy model illustrates critical thermodynamic factors that influence the water/CO 2 splitting capacity. Our research reveals the thermodynamic driving forces underlying the unconventional high-capacity Fe-poor ferrites, further explained via phase diagrams of Fe–Co–O, Fe–Ni–O and Fe–Mg–O. Future materials improvements can be guided by our reduced Gibbs free energy model.

08 HYDROGEN↗
DOE OSTI2024

Leveraging Multitime Hamilton–Jacobi PDEs for Certain Scientific Machine Learning Problems

Hamilton-Jacobi partial differential equations (HJ PDEs) have deep connections with a wide range of fields, including optimal control, differential games, and imaging sciences. By considering the time variable to be a higher dimensional quantity, HJ PDEs can be extended to the multi-time case. In this paper, we establish a novel theoretical connection between specific optimization problems arising in machine learning and the multi-time Hopf formula, which corresponds to a representation of the solution to certain multi-time HJ PDEs. Through this connection, we increase the interpretability of the training process of certain machine learning applications by showing that when we solve these learning problems, we also solve a multi-time HJ PDE and, by extension, its corresponding optimal control problem. As a first exploration of this connection, we develop the relation between the regularized linear regression problem and the Linear Quadratic Regulator (LQR). We then leverage our theoretical connection to adapt standard LQR solvers (namely, those based on the Riccati ordinary differential equations) to design new training approaches for machine learning. Lastly, we provide some numerical examples that demonstrate the versatility and possible computational advantages of our Riccati-based approach in the context of continual learning, post-training calibration, transfer learning, and sparse dynamics identification.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2023

Accurate Prediction of Algal Biomass Lipid, Protein, and Carbohydrate Composition with Machine Learning Regression Modelling of Near-IR Spectra

During large scale algal biomass cultivation, it is difficult to reliably control relative composition to target levels. Rapid determination of chemical composition is feasible by using near infrared (NIR) spectral data. We sought to build and improve on reliable high-throughput screening prediction method based on partial least squares regression (PLSR) by the application of artificial neural networks (ANN) and associated optimization strategies. The algal biomass sample set was designed and created in an iterative process of culturing in physiologically diverse conditions at the GAI field site, followed by compositional analyses at NREL. The workflow allowed us to identify gaps in compositional space for informing the subsequent cultivation and sampling efforts and generated a high quality set of 210 unique samples with chemical analysis results, spectral scanning data, and cultivation metadata. We observed a significant improvement in the performance of carbohydrate content predictions using an optimized ANN model compared to PLSR, with > 16% reduction in mean absolute percent error (MAPE) when tested on the same set of reserved data. The optimized ANN models for FAME and protein prediction performed exceptionally well with 5.99% and 5.09% MAPE, respectively. Application of these methods to detection and quantification of minor biomass constituents that are relevant to certain product streams has shown positive preliminary results, opening the possibility for extensions to the outputs of this powerful data type. All models are accompanied by prediction uncertainties and unsupervised spectral outlier detection to alert an operator to unreliable spectral data. These tools can be deployed for rapid determination of algal culture status, and cultivation and biomass quality improvement.

algal biofuels↗
DOE OSTI2024

Challenges in Training PINNs: A Loss Landscape Perspective

This paper explores challenges in training Physics Informed Neural Networks (PINNs), emphasizing the role of the loss landscape in the training process. We examine difficulties in minimizing the PINN loss function, particularly due to ill conditioning caused by differential operators in the residual term. We compare gradient-based optimizers Adam, L-BFGS, and their combination Adam+L-FGS, showing the superiority of Adam+L-BFGS, and introduce a novel secondorder optimizer, NysNewton-CG (NNCG), which significantly improves PINN performance. Theoretically, our work elucidates the connection between ill-conditioned differential operators and ill-conditioning in the PINN loss and shows the benefits of combining first- and second-order optimization methods. Our work presents valuable insights and more powerful optimization strategies for training PINNs, which could improve the utility of PINNs for solving difficult partial differential equations.

Rathore, Pratik↗
DOE OSTI2021

Optimizing Error-Bounded Lossy Compression for Scientific Data on GPUs

Error-bounded lossy compression is a critical technique for significantly reducing scientific data volumes. With ever-emerging heterogeneous high-performance computing (HPC) architecture, GPU-accelerated error-bounded compressors (such as CUSZ and cuZFP) have been developed. However, they suffer from either low performance or low compression ratios. To this end, we propose CUSZ+ to target both high compression ratios and throughputs. We identify that data sparsity and data smoothness are key factors for high compression throughputs. Our key contributions in this work are fourfold: (1) We propose an efficient compression workflow to adaptively perform run-length encoding and/or variable-length encoding. (2) We derive Lorenzo reconstruction in decompression as multidimensional partial-sum computation and propose a fine-grained Lorenzo reconstruction algorithm for GPU architectures. (3) We carefully optimize each of CUSZ kernels by leveraging state-of-the-art CUDA parallel primitives. (4) We evaluate CUSZ+ using seven real-world HPC application datasets on V100 and A100 GPUs. Experiments show CUSZ+ improves the compression throughputs and ratios by up to 18.4x and 5.3x, respectively, over CUSZ on the tested datasets.

Tian, Jiannan↗
DOE OSTI2021

Electrolytic Reduction of Titanium Dioxide in Molten LiCl–Li2O

The electrolytic reduction of TiO2 in LiCl–Li2O (1 wt.%) at 650 °C was investigated under a series of cathodic reduction potentials and applied charges to provide a mechanistic understanding of the electrochemical characteristics of the system. The optimal cathodic reduction potential was determined as being −0.3 V vs. Li/Li+. Li2TiO3 and LiTiO2 were structurally identified as intermediate and partial reduction products of the TiO2 electrolytic reduction. The reduction of LiTiO2 was extremely slow and reversible due to its high stability and the detrimental effect of Li2O accumulation within the solid particles. The most reduced product obtained in this study was LiTiO2, which was achieved when using 150% of the theoretical charge under the optimal reduction potential. The highest reduction extent obtained in this study was 25%. Based on theoretical DFT modeling, a detailed multistep reduction mechanism and scheme were proposed for TiO2 electrolytic reduction in LiCl–Li2O (1 wt.%) at 650 °C.

Shi, Meng↗
DOE OSTI2022

Inference for Nonparanormal Partial Correlation via Regularized Rank-Based Nodewise Regression

Abstract Partial correlation is a common tool in studying conditional dependence for Gaussian distributed data. However, partial correlation being zero may not be equivalent to conditional independence under non-Gaussian distributions. In this paper, we propose a statistical inference procedure for partial correlations under the high-dimensional nonparanormal (NPN) model where the observed data are normally distributed after certain monotone transformations. The NPN partial correlation is the partial correlation of the normal transformed data under the NPN model, which is a more general measure of conditional dependence. We estimate the NPN partial correlations by regularized nodewise regression based on the empirical ranks of the original data. A multiple testing procedure is proposed to identify the nonzero NPN partial correlations. The proposed method can be carried out by a simple coordinate descent algorithm for lasso optimization. It is easy-to-implement and computationally more efficient compared to the existing methods for estimating NPN graphical models. Theoretical results are developed to show the asymptotic normality of the proposed estimator and to justify the proposed multiple testing procedure. Numerical simulations and a case study on brain imaging data demonstrate the utility of the proposed procedure and evaluate its performance compared to the existing methods. Data used in preparation of this article were obtained from the Alzheimer's Disease Neuroimaging Initiative (ADNI) database.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2023

Optimal Storage Response to Utility Tariff Structures and Potential Use of Capacity Charges

Energy storage is increasingly being deployed in behind-the-meter use cases, partially in response to falling lithium-ion prices and new utility tariff structures. Utilities are grappling with new tariff design in the presence of distributed energy resources (DER), trying to motivate, fairly price the contribution of, and, in some cases, discourage, certain operation of DER. There are opportunities for energy storage under emerging tariff structures, but utility net load management objectives for storage remain unclear. Diverse tariff structures motivate the use of energy storage to provide one or a combination of: energy arbitrage, energy shifting, solar self-consumption, and import shaving. This paper formulates a linear optimization to demonstrate the optimal storage tariff response, examining customer net load metrics under diverse utility tariff structures, such as time-of-use, net metering, feed-in tariffs, zero export tariffs, and demand charges. A key contribution of this paper is the introduction and examination of a capacity charge as mechanism that can both motivate reductions in both peak import, and exports, along with greater load levelling.

behind-the-meter↗
DOE OSTI2021

A data-driven peridynamic continuum model for upscaling molecular dynamics

Nonlocal models, including peridynamics, often use integral operators that embed lengthscales in their definition. However, the integrands in these operators are difficult to define from the data that are typically available for a given physical system, such as laboratory mechanical property tests. In contrast, molecular dynamics (MD) does not require these integrands, but it suffers from computational limitations in the length and time scales it can address. To combine the strengths of both methods and to obtain a coarse-grained, homogenized continuum model that efficiently and accurately captures materials’ behavior, we propose a learning framework to extract, from MD data, an optimal Linear Peridynamic Solid (LPS) model as a surrogate for MD displacements. To maximize the accuracy of the learnt model we allow the peridynamic influence function to be partially negative, while preserving the well-posedness of the resulting model. To achieve this, we provide sufficient well-posedness conditions for discretized LPS models with sign-changing influence functions and develop a constrained optimization algorithm that minimizes the equation residual while enforcing such solvability conditions. This framework guarantees that the resulting model is mathematically well-posed, physically consistent, and that it generalizes well to settings that are different from the ones used during training. We illustrate the efficacy of the proposed approach with several numerical tests for single layer graphene. Our two-dimensional tests show the robustness of the proposed algorithm on validation data sets that include thermal noise, different domain shapes and external loadings, and discretizations substantially different from the ones used for training.

homogenization↗
DOE OSTI2025

Remote Americium Detection Using an Optical Sensor: A D-Optimal Strategy for Efficient PLS-Based Modeling

A fiber-optic visible–near-infrared absorption spectroscopy system in a glove box was demonstrated for remote quantification of Am(III) (0–500 µM) and HNO 3 (0.1–9 M) using partial least squares regression (PLSR) models. The sensor platform, featuring a simple plug-and-play spectrophotometer, can enable noninvasive, real-time monitoring of actinide process solutions. To establish a flexible PLSR model calibration strategy, a D-optimal design developed using Nd(III) in previous studies was successfully extended to an actinide system with Am(III) to effectively minimize sample set size while maintaining robust prediction performance. The results suggest strong spectral similarities between Nd(III) and Am(III) and validate Nd(III) as an effective optical surrogate for trivalent actinide species. This work also supports the generalizability of a D-optimal training set selection approach for two-factor systems. The PLS1 models for Am(III) and HNO 3 outperformed a PLS2 model and maintained reasonable performance in the presence of interfering U(VI). The resulting sensor system and multivariate approach provides a flexible and scalable solution for process monitoring, control, and safety in diverse nuclear applications.

actinide↗
DOE OSTI2025

Modular Processing of Flare Gas for Carbon Nanoproducts

This project demonstrated the technical viability and economic promise of a modular system for converting flared natural gas into valuable carbon nanoproducts (CNPs) through catalytic chemical vapor deposition (CVD). All major project milestones were successfully completed, including reactor design and commissioning, catalyst development, process optimization, technoeconomic analysis, and application testing in concrete systems. The overarching goal was to create a scalable, field-deployable process that valorizes stranded methane by producing high-value carbon materials for use in cementitious composites. At the lab scale, the team designed and built a fluidized bed reactor optimized for use with silica fume-supported nickel catalysts synthesized via atomic layer deposition (ALD). A statistically designed sintering study enabled precise tuning of nickel nanoparticle size, identifying the influence of oxygen partial pressure, time, and temperature on catalyst morphology and performance. These insights allowed the team to target catalyst conditions that maximize carbon nanofilament growth. Subsequent CVD experiments achieved up to 31.8 wt% carbon deposition under optimized conditions, with TEM confirming the presence of nanofilament structures and sustained hydrogen evolution during reaction. Reactor upgrades and empirical fluidization studies supported the development of reliable, repeatable experimental protocols. The modular pilot-scale skid reactor was fully constructed, instrumented, and commissioned. Capable of operating at 675–800°C and pressures up to 290 psig, the system was designed for continuous operation at a carbon production rate of 1 kg/hr. Initial demonstration runs confirmed solids handling, thermal control, and system leak-tightness, although a critical reactor component (the downfeed tube) was inadvertently omitted during final assembly. This omission limited gas–solid contact and prevented meaningful carbon deposition during pilot-scale CVD runs. Nonetheless, the system operated safely under design conditions, and the root cause of performance limitations was clearly identified. Complementary work on UHPC formulations demonstrated that small additions of carbon nanoproducts, including those derived from flare gas, can significantly enhance mechanical performance while preserving workability. A comprehensive study of CNF dispersion techniques and mix design optimization led to a clear protocol for integrating these nanomaterials into concrete. Incorporation of CNPs improved flexural toughness and reduced porosity, supporting their use in high-performance infrastructure applications. A technoeconomic analysis (TEA) confirmed that this process can produce CNP-loaded catalyst material at a levelized cost below $\$$7/kg across a range of catalyst loadings and reaction yields. With estimated market values for the carbon composite product ranging from $\$$14 to over $\$$60/kg, and the ability to blend CNPs into concrete at sub-percent levels with less than 10% added cost, the system presents a compelling economic case. While additional engineering work is needed to optimize fluidization and heat transfer at scale, this project establishes a strong foundation for commercial development. The process is not only technically sound but also economically promising, representing a viable pathway for flare gas mitigation through modular carbon nanomaterial production.

03 NATURAL GAS↗