Search NASA⌕ Search

SEARCH · Search NASA

Results for “performance improvements”

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 OSTI2025

Improving Runtime Performance of Tensor Computations using Rust From Python

In this work, we investigate improving the runtime performance of key computational kernels in the Python Tensor Toolbox (pyttb), a package for analyzing tensor data across a wide variety of applications. Recent runtime performance improvements have been demonstrated using Rust, a compiled language, from Python via extension modules leveraging the Python C API—e.g., web applications, data parsing, data validation, etc. Using this same approach, we study the runtime performance of key tensor kernels of increasing complexity, from simple kernels involving sums of products over data accessed through single and nested loops to more advanced tensor multiplication kernels that are key in low-rank tensor decomposition and tensor regression algorithms. In numerical experiments involving synthetically generated tensor data of various sizes and these tensor kernels, we demonstrate consistent improvements in runtime performance when using Rust from Python over 1) using Python alone, 2) using Python and the Numba just-in-time Python compiler (for loop-based kernels), and 3) using the NumPy Python package for scientific computing (for pyttb kernels).

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2025

Use of Rig Parameter Data in Bit Constraint Models for Improved Drilling Performance at The Geysers

Surface parameter measurements are routinely used during deep well construction to monitor and guide drilling conditions for improved performance and reduced costs. However, these measurements are of reduced value without a standard to aid in evaluation and decision making. A method is demonstrated whereby drill bit constraint models are used to interpret drilling response parameters. Drill rig parameter data for well GDC-36 at the Geysers Geothermal Field Power were acquired by Geysers Power Company and drilling contractor Kenai Drilling using Pason US DataHub and evaluated. Drilling parameters are evaluated using laboratory-validated rock reduction models for predicting the phenomenological response of drag bits (Detournay and Defourny, 1992) along with other model constraints in computational algorithms. The method is used to evaluate overall bit performance, monitor bit integrity, and detect the presence of drillstring vibrations and other conditions contributing to bit failure; comparisons are made to observations of bit wear and damage. The method will be applied in real-time to improve decision-making on subsequent wells and has applicability to development of advanced analytics on future geothermal wells using real-time electronic drilling recorder (EDR) data for improved performance and reduced drilling costs.

15 GEOTHERMAL ENERGY↗
DOE OSTI2025

A Modified DBC Substrate Improving Thermal Performance for Confined Space Applications

High-power modules use substrates to house the semiconductor device and for electrical insulation. These substrates are constructed with thermally conductive dielectric material sandwiched between two metals to extract heat from semiconductor chips. Thus, the required cooling performance of a power module is linked to the substrate’s thermal performance and can vary based on the substrate technologies. Here, in this study, five substrate technologies were evaluated for space-restricted applications: direct-bonded copper (DBC), an insulated metal substrate (IMS), a thermally annealed pyrolytic graphite (TPG)-based IMS, DBC-based double-sided cooling, and direct-bonded aluminum (DBA) where the heat sink is directly attached without thermal interface materials (TIMs). The finite element (FE) analysis results suggest that the popular DBC substrate has thermal performance better than that of the other substrates for space-constrained applications. To further improve the thermal performance, a modified DBC substrate was proposed where a copper block was added between the semiconductor and the DBC substrate to achieve heat spreading underneath the chip. The modified DBC performance was then compared with the aforementioned substrates and the results showed significant thermal performance improvement. The results were verified with experimental results where the proposed substrate showed 20% more loss handling capability compared to an identical DBC substrate.

42 ENGINEERING↗
DOE OSTI2023

GPU-resident sparse direct linear solvers for alternating current optimal power flow analysis

Integrating renewable resources within the transmission grid at a wide scale poses significant challenges for economic dispatch as it requires analysis with more optimization parameters, constraints, and sources of uncertainty. This motivates the investigation of more efficient computational methods, especially those for solving the underlying linear systems, which typically take more than half of the overall computation time. In this paper, we present our work on sparse linear solvers that take advantage of hardware accelerators, such as graphical processing units (GPUs), and improve the overall performance when used within economic dispatch computations. We treat the problems as sparse, which allows for faster execution but also makes the implementation of numerical methods more challenging. We present the first GPU-native sparse direct solver that can execute on both AMD and NVIDIA GPUs. We demonstrate significant performance improvements when using high-performance linear solvers within alternating current optimal power flow (ACOPF) analysis. Furthermore, we demonstrate the feasibility of getting significant performance improvements by executing the entire computation on GPU-based hardware. Finally, we identify outstanding research issues and opportunities for even better utilization of heterogeneous systems, including those equipped with GPUs.

24 POWER TRANSMISSION AND DISTRIBUTION↗
DOE OSTI2023

Performance portable ice-sheet modeling with MALI

High-resolution simulations of polar ice sheets play a crucial role in the ongoing effort to develop more accurate and reliable Earth system models for probabilistic sea-level projections. These simulations often require a massive amount of memory and computation from large supercomputing clusters to provide sufficient accuracy and resolution; therefore, it has become essential to ensure performance on these platforms. Many of today’s supercomputers contain a diverse set of computing architectures and require specific programming interfaces in order to obtain optimal efficiency. In an effort to avoid architecture-specific programming and maintain productivity across platforms, the ice-sheet modeling code known as MPAS-Albany Land Ice (MALI) uses high-level abstractions to integrate Trilinos libraries and the Kokkos programming model for performance portable code across a variety of different architectures. In this article, we analyze the performance portable features of MALI via a performance analysis on current CPU-based and GPU-based supercomputers. The analysis highlights not only the performance portable improvements made in finite element assembly and multigrid preconditioning within MALI with speedups between 1.26 and 1.82x across CPU and GPU architectures but also identifies the need to further improve performance in software coupling and preconditioning on GPUs. We perform a weak scalability study and show that simulations on GPU-based machines perform 1.24–1.92x faster when utilizing the GPUs. The best performance is found in finite element assembly, which achieved a speedup of up to 8.65x and a weak scaling efficiency of 82.6% with GPUs. We additionally describe an automated performance testing framework developed for this code base using a changepoint detection method. The framework is used to make actionable decisions about performance within MALI. We provide several concrete examples of scenarios in which the framework has identified performance regressions, improvements, and algorithm differences over the course of 2 years of development.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2023

Improved Primary Reference Cell Calibrations for Higher Accuracy Photovoltaic Cell and Module Performance Measurements

The adoption of photovoltaic (PV) modules for clean electricity relies on accurate measurements of their performance, which are essential for estimating their energy production potential. Herein, the calibration chain of PV cells and modules, with particular emphasis on primary reference cell calibrations, is discussed. Also, herein, the direct sunlight method the group has developed for these calibrations is presented and critical improvements and upgrades that lead to calibration uncertainty as low as 0.45% are discussed. The ultimate motivation behind this work is to provide low‐uncertainty performance measurements of PV modules, and lowering the calibration uncertainty of primary reference cells is a key first step toward achieving this goal. As the use of solar electricity continues to grow, the demand for primary reference cell calibrations inevitably increases beyond what the small handful of primary calibration laboratories can provide today. Therefore, this work can serve as a useful guide for implementing primary PV reference cell calibrations using the outdoor method, as well as outlining the critical elements required to make these calibrations highly accurate.

Osterwald, Carl R.↗
DOE OSTI2024

Development of Stable Solid Oxide Electrolysis Cells for Low-Cost Hydrogen Production

The project objective was to demonstrate a solid oxide cell-based steam electrolysis stack that exhibits robustness, reliability, endurance, hydrogen purity, and produces hydrogen at elevated pressure of 2 to 3 bar. Innovative materials and processing methods were evaluated to improve degradation characteristics. Performance improvement focused on nearly all layers involved in the cell and stack assembly. Primary attention was paid to zirconia-ceria interface resistance control via sintering optimization and decrease in degradation from the oxygen electrode by evaluating low strontium (Sr) or Sr-free composition for both the oxygen electrode and current collection layer. Stack robustness was addressed by validating redox tolerance of fuel electrode, confirming capability of cells to survive repeated thermal cycles, studying the effect of pressure on performance and degradation, evaluating the effect of contamination on fuel and oxygen electrode performance and degradation, and identifying mitigation strategies to improve performance. The characterization included evaluation of electrochemical performance and stability followed by microstructural analysis. At the cell level, performance and stability improvements were achieved by incorporating a Sr-free oxygen electrode and a denser oxygen electrode barrier layer. At the stack level, pressurized operation reduces demand on first stage compression, the redox tolerant fuel electrode mitigates risk from service interruptions, and improvements to interconnect coating alleviate chromium (Cr) contamination effects. The denser barrier layer was achieved by adding a sintering aid to the samaria-doped ceria (SDC) composition that reduced sintering temperature by 150 °C. The resulting density was on par with the baseline SDC barrier layer density and the lower sintering temperature resulted in less resistive phase formation during sintering. Button cell tests did not demonstrate a change in performance when exposed to silicon (Si) or manganese (Mn) impurities to the fuel electrode and Cr impurity to the oxygen electrode. More detailed study however is warranted. The project addressed SOEC performance and stability at the cell and stack levels through a systematic approach to known sources of degradation that were combined and tested in three stack tests using an electrolyte supported cell design to allow for evaluation of a variety of fuel and oxygen electrode compositions. STK-82 and STK-83 had identical compositions. STK-100 incorporated the best materials and processing variables developed under this and concurrent projects, and was tested at elevated pressure in steam electrolysis. • STK-82 recovered performance after redox and thermal cycling, demonstrating the robustness of the stack and seals. It exhibited stable performance in testing for 500 hours in SOEC mode, followed by 300 hours of cycling between SOEC and SOFC tests. Degradation during SOEC operation was 1.8 %/ 1,000 hours. • STK-83 generated hydrogen at >80% steam conversion, and oxygen above 98.5 % purity during pressurized operation. Both hydrogen and oxygen were generated at 3 barg pressure without the use of a pressure vessel. In addition to balanced pressure, electrolysis operation at 1 bar differential pressure across anode and cathode was also demonstrated to substantial the robustness of the cell and seal. • STK-100 measured at initial ambient pressure conditions showed an area specific resistance of 1.1 ohm-cm 2 , and STK-83 had 1.3 ohm-cm 2 .

08 HYDROGEN↗
DOE OSTI2020

Improving the Performance of Integrated Power-Hardware-in-the-Loop and Quasi-Static Time-Series Simulations

This paper introduces a novel multirate cosimulation architecture that overcomes previous challenges integrating faster real-time, microseconds-scale, power-hardware-in-the-loop (PHIL) with larger scale but slower, real-time, seconds-scale, quasi-static time-series (QSTS) simulation. Specifically, an intermediate reduced-equivalent electromagnetic transient (EMT) model duplicates the key power system topology to capture high-speed dynamics and converge hardware-power system interactions between QSTS updates. Exchanging multidimensional parameter vectors (loads, control status, etc.) between the QSTS and EMT models enables capturing the interactions of the rich, high-node-count (thousands of electrical nodes) QSTS model with the hardware. This architecture offers full spatial resolution from QSTS, including individual load dynamics, actual distribution management system controls, and nodal voltages. Simultaneously, the intermediate EMT model provides more detailed high-speed transients of key distributed energy resource hardware interactions. In addition, we use careful PHIL interface design and the exchange of complex power data, rather than current, to further improve performance. This cosimulation architecture is demonstrated by testing a simulated distribution system with an interconnected 500-kVA advanced photovoltaic (PV) inverter in PHIL. In conclusion, the architecture successfully captured PV local volt-volt ampere reactive (volt-VAR) control interactions with the larger network under time-varying electrical and weather conditions.

42 ENGINEERING↗
DOE OSTI2023

High-Nickel Cathode Materials for High-Energy, Long-Life, Low-Cost Lithium-Ion Batteries (Final Report)

The overall goal of this project is to produce Li-ion cells, which deliver at least 600 Wh kg -1 at the cathode level with usable lifetimes of at least 1,000 cycles or 15 years of storage (calendar life). A secondary goal of this project is to minimize the use of cobalt in the cathode, with the ultimate goal of eliminating cobalt entirely. To develop such a cathode, the University of Texas at Austin (UT Austin) team extensively investigated the structural, physical, and chemical properties of layered oxide materials. A panel of dopant elements were surveyed with an emphasis on identifying those, which impart electrochemical and thermal stability. Of those tested, Al and Mg showed the greatest performance improvements while still being easily incorporated into the cathode through simple alterations of the precursor coprecipitation method. Other dopants, such as Ta, Nb, B, and F significantly improved cathode performance, yet their mode of incorporation required additional processing which negated their use under scaled-up conditions. From this work, a first-generation cathode LiNi 0.85 Mn 0.075 Co 0.05 Al 0.02 Mg 0.005 O 2 (NMCAM-85) was produced and an initial set of 2 Ah cells was fabricated by Tesla Inc. These cells delivered 650 Wh kg -1 and a cycle life of just over 700 cycles. Further improvements were garnered through careful study of electrolyte additives. A tailored, ethylene carbonate (EC)-free electrolyte was developed, which improved capacity retention from 60% to 82% after 500 cycles in model LiNi 0.9 Mn 0.05 Al 0.05 O 2 (NMA-90) full cells. This study identified the additives fluoroethylene carbonate (FEC) and lithium difluorophosphate (LiPO 2 F 2 ) as beneficial additives, which were included in the electrolyte for the end-of-project cells. The UT Austin team developed a high-Ni layered oxide cathode LiNi 0.85 Mn 0.1 Al 0.05 Mg 0.005 O 2 (NMAM-85) capable of meeting the project goals. This cathode was synthesized in kilogram quantities and delivered to Tesla for fabrication into 2 Ah pouch cells for DOE testing. These large-format cells achieved 630 Wh kg -1 at the cathode level under the designated testing parameters, meeting the project goal with a completely cobalt-free cathode. The large-format cells are currently undergoing cycle life and calendar life evaluation at Idaho National Lab.

25 ENERGY STORAGE↗
DOE OSTI2020

Mathematical Modeling of Hydroxide-Exchange-Membrane Water Electrolyzer

Water electrolyzers can transform intermittent renewable energy like solar energy and wind energy into the chemical energy of hydrogen with zero greenhouse-gas emissions. The hydroxide-exchange membrane electrolyzer (HEME) combines the capability to produce pressurized hydrogen with the advantage of being able to use low or non-platinum group metal (PGM) electrocatalysts in the alkaline environment.1 Hydroxide salts, for example, KOH, are added to the HEME water feed on both anode and cathode to improve its performance. However, the specific mechanism of performance improvement still needs to be further understood. In addition, at high current densities, bubble evolution can result in mass-transport limitations, a less well studied phenomena. Mathematical modeling is ideal to explore these issues as it is cost and time efficient and can deconvolute the physics, processes, and observed phenomena and study the applied-voltage breakdown. In this work, we extend our previously developed 1D two-phase continuum model2 to study the varies processes in the HEME and provide insights on performance optimizations. First, the oxygen evolution reaction (OER) and hydrogen evolution reaction (HER) kinetics at different hydroxide concentrations have been studied by rotating disk electrodes (RDE) and implemented in the model. Then, the model is calibrated and validated against experimental HEME polarization curves for different KOH concentrations as a liquid electrolyte. The model clearly shows a performance increase with increasing KOH concentrations, which is consistent with the experimental results. The reduced ohmic resistance and increased electrochemical active surface area (ECSA) are the two main reasons for performance increase. The large amount of hydroxide in the liquid electrolyte not only helps to distribute the reactant hydroxide throughout the catalyst layer (CL), which reduces ohmic loss, but also enables reaction at the interface between the liquid electrolyte and electrocatalyst, which increases the ECSA. Applied-voltage breakdown demonstrates that the electrolyzer performance is dominated by anode kinetics and ohmic loss. A comparison with the DI water feed shows a more uniform current distribution in the anode CL when KOH is added, which indicates a higher utilization of the CL. Second, we present modeling on the effects of bubble coverage. As gas evolves, part of the ECSA is minimized due to bubble coverage. To account for this effect, an empirical relationship between the fractional bubble coverage and the current density is implemented in the model.3 The model shows this bubble coverage effect is more pronounced at large current densities with DI water feed. Acknowledgements This work was funded under the HydroGEN Consortium by the Energy Efficiency and Renewable Energy, Hydrogen and Fuel Cell Technologies Office, of the U. S. Department of Energy under contract number DE-AC02-05CH11231. References R. Abbasi, B. P. Setzler, S. Lin, J. Wang, Y. Zhao, H. Xu, B. Pivovar, B. Tian, X. Chen, G. Wu and Y. Yan, 31, 1805876 (2019). L. N. Stanislaw, M. R. Gerhardt and A. Z. Weber, ECS Transactions, 92, 767 (2019). H. Vogt and R. J. Balzer, Electrochimica Acta, 50, 2073 (2005).

Liu, Jiangjin↗
DOE OSTI2020

Optimizing the hypre solver for manycore and GPU architectures

The solution of large-scale combustion problems with codes such as Uintah on modern computer architectures requires the use of multithreading and GPUs to achieve performance. Uintah uses a low-Mach number approximation that requires iteratively solving a large system of linear equations. The Hypre iterative solver has solved such systems in a scalable way for Uintah, but the use of OpenMP with Hypre leads to at least slowdown due to OpenMP overheads. The proposed solution uses the MPI Endpoints within Hypre, where each team of threads acts as a different MPI rank. This approach minimizes OpenMP synchronization overhead and performs as fast or (up to 1.44) faster than Hypre's MPI-only version, and allows the rest of Uintah to be optimized using OpenMP. The profiling of the GPU version of Hypre shows the bottleneck to be the launch overhead of thousands of micro-kernels. The GPU performance was improved by fusing these micro-kernels and was further optimized by using Cuda-aware MPI, resulting in an overall speedup of 1.16—1.44 compared to the baseline GPU implementation. The above optimization strategies were published in the International Conference on Computational Science 2020 [1]. This work extends the previously published research by carrying out the second phase of communication-centered optimizations in Hypre to improve its scalability on large-scale supercomputers. Additionally, this includes an efficient non-blocking inter-thread communication scheme, communication-reducing patch assignment, and expression of logical communication parallelism to a new version of the MPICH library that utilizes the underlying network parallelism [2]. The above optimizations avoid communication bottlenecks previously observed during strong scaling and improve performance by up to 2 on 256 nodes of Intel Knight's Landing processor.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2022

Posterior Regularized Bayesian Neural Network incorporating soft and hard knowledge constraints

Neural Networks (NNs) have been widely used in supervised learning due to their ability to model complex nonlinear patterns, often presented in high-dimensional data such as images and text. However, traditional NNs often lack the ability for uncertainty quantification. Bayesian NNs (BNNS) could help measure the uncertainty by considering the distributions of the NN model parameters. Besides, domain knowledge is commonly available and could improve the performance of BNNs if it can be appropriately incorporated. In this work, we propose a novel Posterior-Regularized Bayesian Neural Network (PR-BNN) model by incorporating different types of knowledge constraints, such as the soft and hard constraints, as a posterior regularization term. Furthermore, we propose to combine the augmented Lagrangian method and the existing BNN solvers for efficient inference. Furthermore, the experiments in simulation and two case studies about aviation landing prediction and solar energy output prediction have shown the knowledge constraints and the performance improvement of the proposed model over traditional BNNs without the constraints.

14 SOLAR ENERGY↗
DOE OSTI2022

Prospects for additive manufacturing of nuclear fuel forms

Additive manufacturing (AM) is viewed as a critical enabling technology for achieving superior performance and improved economics across many sectors. Although the past decade has seen increased application of AM methods to nuclear reactor cladding and structural materials, exploration of AM as applied to the uranium-bearing nuclear fuel forms has been limited. The two major families of nuclear fuel forms (monolithic, particle/dispersion) are utilized with their own set of core objectives in mind, and their differences require distinct fabrication infrastructures. These reference fabrication methods impose many limitations on nuclear fuels. Further, both currently operating reactors and future concepts have the potential for improved performance if these accepted limitations are relaxed or removed entirely. The primary limitations of reference fabrication processes for the common nuclear fuel forms are outlined in this paper. This groundwork is then used to identify avenues of fuel performance specific to each of these fuel architectures that could be exploited if the restrictions of conventional fuel fabrication are removed. Moreover, multiple targets for AM studies are laid out for each of the major nuclear fuel variants. Finally, key strategic components to guide research activities in AM of nuclear fuels are outlined, with an emphasis on use of modeling and simulation to motivate research aims and embrace of an accelerated testing methodology to screen and quality new fuel forms.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗
DOE OSTI2020

Thermionic microwave gun for terahertz and synchrotron light sources

Conventional thermionic microwave and radiofrequency (RF) guns can offer high average beam current, which is important for synchrotron light and terahertz (THz) radiation source facilities, as well as for industrial applications. For example, the Advanced Photon Source at Argonne National Laboratory is a national synchrotron radiation light source research facility that utilizes thermionic RF guns. Yet, these existing thermionic guns are bulky, difficult to handle and install, easily detuned, very sensitive to thermal expansion, and are due for a major upgrade and replacement. Here, we present the design of a new, more stable and reliable gun with optimized electromagnetic performance, improved thermal engineering and a more robust cathode mounting technique, which is a critical step to improve the performance of existing and future light sources, industrial accelerators, and electron beam-driven terahertz sources. We will also present a fabricated gun prototype and show results of high-power and beam tests.

47 OTHER INSTRUMENTATION↗
DOE OSTI2022

SchedInspector: A Batch Job Scheduling Inspector Using Reinforcement Learning

Improving the performance of job executions is an important goal of HPC batch job schedulers, such as minimizing job waiting time, slowdown, or completion time. Such a goal is often accomplished using carefully designed heuristics based on job features, such as job size and job duration. However, these heuristics overlook important runtime factors (e.g., cluster availability and waiting job patterns), which may vary across time and make a previously sound scheduling decision not hold any longer. In this study, we propose a new approach to incorporate runtime factors into batch job scheduling for better job execution performance. The key idea is to add a scheduling inspector on top of the base job scheduler to scrutinize its scheduling decisions. The inspector will take the runtime factors into consideration and accordingly determine the fitness of the scheduled job. It then either accepts the scheduled job or rejects it and asks the base schedulers to try again later. We realize such an inspector, namely SchedInspector, by leveraging the intelligence of reinforcement learning. Through extensive experiments, we show SchedInspector can intelligently integrate the runtime factors into various batch job scheduling policies, including the state-of-the-art one, to gain better job execution performance, such as smaller average bounded job slowdown (up to 69% better) or average job waiting time (up to 52% better), across various real-world workloads. We also show that although rejecting scheduling decisions may leave the resources idle hence affect the system utilization, SchedInspector is able to achieve the job execution performance improvement with marginal impact on the system utilization (typically less than 1%). We consider one key advantage of SchedInspector is it automatically learns to work with and improve existing job scheduling policies without changing them, which makes it promising to serve as a generic enhancer for various batch job scheduling policies.

Zhang, Di↗
DOE OSTI2022

Posterior Regularized Bayesian Neural Network

Traditional NNs often lack the ability for uncertainty quantification. Bayesian NNs(BNNs) could help measure the confidence level by using distributions in NNs modeling. Besides, knowledge is commonly available and could improve the performance of BNNs if it can be properly incorporated. In this work, we propose a novel Posterior-Regularized BNN(PR-BNN) model by incorporating soft and hard constraints as a posterior regularization term. We also propose an augmented Lagrangian method and stochastic optimization algorithm for efficient updating via Monte Carlo sampling. The simulations and case studies for solar PV plants have shown the performance improvement of the proposed model over traditional BNNs.

97 MATHEMATICS AND COMPUTING↗