SEARCH · Engineering Papers
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.
Sparse coding has a variety of applications, such as image denoising and data compression. Recent work has shown the how sparse coding and biologically inspired architectures can mitigate adversarial attacks on images.
The Lab’s advent of the atomic bomb helped end World War II in 1945 and also inspired an entire genre of pop culture, including movies, music, and comic books.
The scientifically challenging problem of catalytically reducing oxyanions was explored and presented great opportunities in terms of the environment, economics and energy self-sufficiency. Oxyanions are pervasive as they are found in many areas of technology, all forms of life, in minerals, and as synthetic materials. Given their ubiquity, the contamination of inorganic oxyanions in drinking water is a national problem as 26 states and Puerto Rico have reported high concentrations of these harmful pollutants. The goals achieved by this research was to catalytically reduce oxyanions utilizing sustainable earth abundant catalysts featuring bio-inspired ligands. This worked allowed us to gain fundamental insights into what dictates the reduction/reactivity for these anions.
With increasing simulation and measurement data, machine learning and artificial intelligence have been widely used in computational decision-making of complex engineering systems. The resulting tools, such as uncertainty quantification solvers, reinforcement learning, and physics-informed machine learning, have achieved great success in critical DOE tasks such as material discovery and design, energy system modeling and control, and numerical weather and climate prediction. A core topic in scientific machine learning and artificial intelligence is Bayesian inference: given an observed data set, people want to estimate the posterior distribution of a (possibly large) number of hidden parameters. Due to the flexibility and weak assumptions, Bayesian sampling has been the mainstream Bayesian inference solvers despite the rapid progress of approximate Bayesian inference. Classical Bayesian sampling methods such as Markov-chain Monte Carlo suffer from a low-acceptance rate due to the random walk nature, therefore state-of-the-art techniques use Hamiltonian Monte Carlo and its variants to efficiently draw posterior samples in a high dimension. The key idea of Hamiltonian Monte Carlo and its variants is to simulate the Hamiltonian dynamics of a classical particle with a fixed mass, and their performance significantly degrades when the posterior distribution is highly spiky or has multiple modes. Leveraging the idea of quantum physics, this project has investigated new theory, algorithms and applications of Bayesian inference (especially Bayesian sampling). The main results include: (1) novel quantum-inspired Bayesian sampling methods that can lead to better accuracy for challenging multi-modal or spiky distributions, (2) more scalable machine learning framework leveraging tensor-compressed Bayesian inference, and (3) Bayesian and sampling approaches for verifying the robustness of continuous and binary neural networks.
Thermal effects of wing color for Albatross-inspired drones performing in the Martian atmosphere are investigated during the summer and winter seasons. This study focuses on two useful consequences of the thermal effects of wing color: the drag reduction and the thermoelectric generation of power. According to its color, each wing side has a certain temperature affecting the drag. Investigations of various configurations have shown that the thermal effect on the wing boundary layer skin drag is insignificant because of the low atmospheric pressure. However, the total drag varies as much as 12.8% between the highest performing wing color configuration and the lowest performing configuration. Additionally, the large temperature differences between the top and the bottom wing surfaces show great potential for thermoelectric power generation. The maximum temperature differences between the top and bottom surfaces for the summer and winter seasons are, respectively, 65 K and 30 K. The drag reduction and the power generation via thermoelectric generators both contribute to enhancing the endurance of drones. Future drone designs will benefit from increased endurance through optimizing the wing color configuration.
We explore and enhance the ability of neu- ral language models to generate novel scien- tific directions grounded in literature. Work on literature-based hypothesis generation has traditionally focused on binary link prediction— severely limiting the expressivity of hypothe- ses. This line of work also does not focus on optimizing novelty. We take a dramatic depar- ture with a novel setting in which models use as input background contexts (e.g., problems, experimental settings, goals), and output natu- ral language ideas grounded in literature. We present SCIMON, a modeling framework that uses retrieval of “inspirations” from past scien- tific papers, and explicitly optimizes for novelty by iteratively comparing to prior papers and up- dating idea suggestions until sufficient novelty is achieved. Comprehensive evaluations reveal that GPT-4 tends to generate ideas with over- all low technical depth and novelty, while our methods partially mitigate this issue. Our work represents a first step toward evaluating and developing language models that generate new ideas derived from the scientific literature.
Abstract. There is growing interest in data-driven weather prediction (DDWP), e.g., using convolutional neural networks such as U-NET that are trained on data from models or reanalysis. Here, we propose three components, inspired by physics, to integrate with commonly used DDWP models in order to improve their forecast accuracy. These components are (1) a deep spatial transformer added to the latent space of U-NET to capture rotation and scaling transformation in the latent space for spatiotemporal data, (2) a data-assimilation (DA) algorithm to ingest noisy observations and improve the initial conditions for next forecasts, and (3) a multi-time-step algorithm, which combines forecasts from DDWP models with different time steps through DA, improving the accuracy of forecasts at short intervals. To show the benefit and feasibility of each component, we use geopotential height at 500 hPa (Z500) from ERA5 reanalysis and examine the short-term forecast accuracy of specific setups of the DDWP framework. Results show that the spatial-transformer-based U-NET (U-STN) clearly outperforms the U-NET, e.g., improving the forecast skill by 45 %. Using a sigma-point ensemble Kalman (SPEnKF) algorithm for DA and U-STN as the forward model, we show that stable, accurate DA cycles are achieved even with high observation noise. This DDWP+DA framework substantially benefits from large (O(1000)) ensembles that are inexpensively generated with the data-driven forward model in each DA cycle. The multi-time-step DDWP+DA framework also shows promise; for example, it reduces the average error by factors of 2–3. These results show the benefits and feasibility of these three components, which are flexible and can be used in a variety of DDWP setups. Furthermore, while here we focus on weather forecasting, the three components can be readily adopted for other parts of the Earth system, such as ocean and land, for which there is a rapid growth of data and need for forecast and assimilation.
Renewable water power, including hydropower and marine energy, will play a key role in building a reliable and flexible 100% clean energy future. That future needs a larger, modern workforce - one that's more diverse, equitable, and inclusive - to power and improve these technologies. And researchers at the National Renewable Energy Laboratory (NREL) are committed to fostering tomorrow's water power workforce through science, technology, engineering, and mathematics (STEM) and workforce development programs. Through events, online resources, and more, the lab aims to engage and inspire students to dive into careers in water power.
Monitoring cutting forces for process control may be challenging because force measurements typically require invasive instrumentation. To remedy this situation, two new methods were recently developed to estimate cutting forces in real time based on the use of on-machine accelerometer measurements. One method uses machine learning, while another uses a physics-inspired data-driven approach, to generate a model that estimates cutting forces from on-machine accelerations. The estimated forces from both approaches were compared against cutting force data collected during various milling operations on several machine tools. The results reveal the advantages and disadvantages of each model to estimate real-time cutting forces.
Renewable water power, including hydropower and marine energy, will play a key role in building a reliable and flexible 100% clean energy future. That future needs a larger, modern workforce - one that's more diverse, equitable, and inclusive - to power and improve these technologies. And researchers at the National Renewable Energy Laboratory (NREL) are committed to fostering tomorrow's water power workforce through science, technology, engineering, and mathematics (STEM) and workforce development programs. Through events, online resources, and more, the lab aims to engage and inspire careers in water power.
Co-locating solar projects and agriculture can provide mutual benefits to local farmers (e.g., dual revenue streams, increased yields from pollinator services, irrigation reductions) and to the solar projects (e.g., reduced timeline/costs for installation and O&M, expanded market, increased PV efficiency from a cooler, vegetated microclimate). While prior work sought to demonstrate the feasibility of agricultural co-location (or "agrivoltaic") opportunities, there is a fundamental gap in data available to developers, landowners, and state agencies that prevents widespread deployment of these mutually beneficial practices. This project addressed this gap by (1) establishing a stakeholder research working group composed of multi-sector industry leaders and academics to provide guidance and assist with outreach; (2) undertaking targeted field-based research projects evaluating solar and agriculture co-location tradeoffs; (3) conducting analysis and modeling studies that complement the field-based research; and (4) developing a data portal to consolidate all data on this topic in one location.
Self-assembly of 3D structures introduce an attractive and scalable route to realize reconfigurable and functionally capable mesoscale devices without human intervention. A common approach for achieving this is to utilize stimuli-responsive folding of hinged structures, which requires the integration of different materials and/or geometric arrangements along the hinges. It is also demonstrated that the inclusion of Kirigami cuts in planar, hingeless bilayer thin sheets can be used to produce complex 3D shapes in an on-demand manner. Nonlinear finite element models are developed to elucidate the mechanics of shape morphing in bilayer thin sheets and verify the predictions through swelling experiments of planar, millimeter-scaled PDMS (polydimethylsiloxane) bilayers in organic solvents. Building upon the mechanistic understandings, The transformation of Kirigami-cut simple bilayers into 3D shapes such as letters from the Roman alphabet (to make “ADVANCED FUNCTIONAL MATERIALS”) and open/closed polyhedral architectures is experimentally demonstrated. A possible application of the bilayers as tether-less optical metamaterials with dynamically tunable light transmission and reflection behaviors is also shown. As the proposed mechanistic design principles could be applied to a variety of materials, this research broadly contributes toward the development of smart, tetherless, and reconfigurable multifunctional systems.
The dream to produce green, clean and sustainable hydrogen from earth-abundant and free resources such as seawater and sunlight is highly motivating because of the interest for desirable economical and societal applications in energy. However, it remains challenging to develop an efficient and unassisted photocatalytic device to split seawater molecules with just sunlight without any external bias. For the first time, a such novel hierarchical material has been developed, based on thin film technology that integrates TiO 2 semiconductor layers, embedded gold nanoparticles and a photosensitive carbo-benzene layer. The design of the triptych device placing the photocatalyst in be-to-be increases the photoactive surface area by a factor 2. Its stability (>120 successive hours of hydrogen production and >5 days in a day/night representative alternation) and efficiency (STH 0.06%) are measured in NaCl-salted water. Here, the chloride ions act as hole scavengers and induce an increase of pH increasing in turn the sun-driven hydrogen production rate.
Foaming prediction is critical for selecting materials and designing processes in industries such as bioprocessing and gas processing. Existing models lack the generality needed for a wide range of materials and overlook the foaming behavior in pure liquids. Here, this work presents a novel method for predicting foaming in pure liquids based on their density, surface tension, and viscosity, using Reynolds ( Re ) and Ohnesorge ( Oh ) numbers. A foaming prediction map, leveraging the theory of fluid drop behavior, was developed by plotting these numbers. This map delineates distinct non-foaming and foaming regions, functioning as a binary classifier for foaming predictions. The map was fitted and validated through shake test experiments on 46 liquids, demonstrating reliable predictions, except for a specific region characterized by small Oh and large Re numbers. This region corresponded to relatively low foam stability and high turbulence, making foaming predictions challenging for liquids in this category.
Thermally adaptive textiles (TATs) enable human subjects to attain thermal comfort without energy consumption, which can lead to enormous energy savings on heating, ventilation, and air conditioning (HVAC) in buildings. Herein, TAT structures which respond to the sweat and generate pores by opening an array of flap‐shaped pores patterned on the fabric surface are proposed. A moisture‐driven self‐actuator for flap opening by constructing a bilayer consisting of a hygroscopic layer using polyethylene glycol and cellulose acetate, and a hydrophobic polymer using a polyester type polymer, is used and successfully demonstrated an essentially instant 4 °C apparent temperature cooling performance within one minute of sweat–humidity‐initiated actuation while wearing TAT using a sweating skin simulated device.
Abstract The organic ligand environment surrounding enzymatic and homogeneous catalytic active sites often determines catalytic activity. Ruthenium nanoparticles, ≤1 nm in diameter, are synthesized using monodentate thiol, monodentate phosphine, and bidentate bisphosphine ligands. Even though some of the ruthenium surface is blocked by the ligands, catalytic activity is still observed for CO oxidation and H 2 O 2 decomposition. All three ligand‐stabilized ruthenium nanoparticles have similar CO oxidation rates; however, the bisphosphine‐stabilized Ru nanoparticles are approximately 2.5 times less active than the monothiol‐stabilized and monophosphine‐stabilized ruthenium nanoparticles for H 2 O 2 decomposition. It is observed that the organic ligand environment is modulated in situ during nanoparticle synthesis via partial oxidation of the bisphosphine as confirmed by 31 P NMR measurements. We hypothesize that bisphosphine‐bound Ru nanoparticles consist of a Ru core with some of the ligands bound in a monodentate manner where the other P atom is oxidized and not bound to the Ru surface leading to a thicker hydrophobic layer around the Ru nanoparticles. The increase in hydrophobicity is confirmed via contact angle and zeta potential measurements. H 2 O 2 decomposition rates are known to decrease with increasing hydrophobicity, and this work illustrates a pathway for increasing hydrophobicity in situ using ligand‐bound metallic nanoparticles.
Designing model-independent anomaly detection algorithms for analyzing LHC data remains a central challenge in the search for new physics, due to the high dimensionality of collider events. In this work, we develop a graph autoencoder as an unsupervised, model-agnostic tool for anomaly detection, using the LHC Olympics dataset as a benchmark. By representing jet constituents as a graph, we introduce a method to systematically control the information available to the model through sparse graph constructions that serve as physically motivated inductive biases. Specifically, (1) we construct graph autoencoders based on locally rigid Laman graphs and globally rigid unique graphs, and (2) we explore the clustering of jet constituents into subjets to interpolate between high- and low-level input representations. We obtain the best performance, measured in terms of the Significance Improvement Characteristic curve for an intermediate level of subjet clustering and certain sparse unique graph constructions. We further investigate the role of graph connectivity in jet classification tasks. Our results demonstrate the potential of leveraging graph-theoretic insights to refine and increase the interpretability of machine learning tools for collider experiments.