Search NASA⌕ Search

SEARCH · Search NASA

Results for “physical 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 1,765 records · Page 98

Storing Co 2 in Built Infrastructure: Co 2 Carbonation of Precast Concrete Products

The overall objective of the proposed study was to advance the technical understanding of CO 2 incorporation into novel cementitious materials for the development of high value products that provide a net reduction in carbon emissions. This project combined two primary phases of research that addressed technical barriers related to (i) optimizing CO 2 storage capability of cementitious materials, (ii) evaluating and enhancing physical properties of novel carbonated materials, and (iii) assessing the reductions in life cycle CO 2 emissions attributed to CO 2 carbonation of precast cementitious materials. Engineered cementitious composites (ECC) are a class of highly ductile concrete composites that have been shown to be very durable when used in the built environment. CO 2 carbonation of ECC was examined in this study and it was found that precast ECC specimens could sequester up to 35% CO 2 by cement mass after 24 hours of curing at a CO 2 pressure of 0.5 MPa and 23°C and had a strain capacity of 3%. Carbonation conditions were optimized at the bench-scale and then utilized to create full-scale CO 2 -cured ECC railroad ties that were field tested on a train track. Rail ties were selected for this initial assessment of CO 2 storage in precast concrete materials due to the large market for concrete ties in the railroad industry. Although the full-scale rail ties passed all of the required American Railway Engineering and Maintenance-of-Way Association qualifying mechanical tests, on-track testing of the CO 2 -cured ECC rail ties was unsuccessful due to fiber alignment in the ECC the during the rail tie casting process which prevented the material from achieving the expected level of strain capacity. This result highlights the challenge in scaling up bench-scale processes to full-scale product manufacturing and requires additional investigation into the casting process of large-scale infrastructure elements using ECC combined with carbonation curing. Life cycle assessment of a CO 2 -cured ECC rail tie versus a traditional concrete rail tie indicates that the ECC tie can have lifecycle carbon savings of between 11% and 51% depending on how much longer its useful lifetime is compared to traditional concrete rail ties. Both carbon and cost savings are driven by a reduction in the need to replace broken rail ties, so the key factor is the extent to which a CO 2 -cured ECC rail tie will have increased lifetime durability compared to alternative rail ties.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

Active learning of ternary alloy structures and energies

Abstract Machine learning models with uncertainty quantification have recently emerged as attractive tools to accelerate the navigation of catalyst design spaces in a data-efficient manner. Here, we combine active learning with a dropout graph convolutional network (dGCN) as a surrogate model to explore the complex materials space of high-entropy alloys (HEAs). We train the dGCN on the formation energies of disordered binary alloy structures in the Pd-Pt-Sn ternary alloy system and improve predictions on ternary structures by performing reduced optimization of the formation free energy, the target property that determines HEA stability, over ensembles of ternary structures constructed based on two coordinate systems: (a) a physics-informed ternary composition space, and (b) data-driven coordinates discovered by the Diffusion Maps manifold learning scheme. Both reduced optimization techniques improve predictions of the formation free energy in the ternary alloy space with a significantly reduced number of DFT calculations compared to a high-fidelity model. The physics-based scheme converges to the target property in a manner akin to a depth-first strategy, whereas the data-driven scheme appears more akin to a breadth-first approach. Both sampling schemes, coupled with our acquisition function, successfully exploit a database of DFT-calculated binary alloy structures and energies, augmented with a relatively small number of ternary alloy calculations, to identify stable ternary HEA compositions and structures. This generalized framework can be extended to incorporate more complex bulk and surface structural motifs, and the results demonstrate that significant dimensionality reduction is possible in thermodynamic sampling problems when suitable active learning schemes are employed.

Chemistry↗

Using Artificial Neural Networks to Predict Physical Properties of Membrane Polymers

Membrane polymers are a promising technology for use in many challenging gas separation applications. Here, the techniques of computer-aided molecular design can be used to search through the massive molecular space of heteropolymers and develop a set of likely candidate repeat units matching specific physical property targets. However, reasonably accurate property prediction algorithms are needed, but these algorithms must be very fast in order to be combined with an optimization framework. Artificial neural networks (ANNs), a branch of machine learning, are applied in this work to predict the physical properties of polymers. All of the physical properties investigated were found to be predicted by ANNs with R 2 scores exceeding 0.82.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

3D reconstruction and neural rendering for adversarial machine learning

While evasion attacks on computer vision systems have been widely studied, creating attacks that remain effective under significant changes in viewpoint continues to be challenging. Traditional approaches often rely on affine transformations of images, but these approaches degrade at larger perspective shifts and often produce unrealistic or ineffective perturbations. Recent methods use differentiable renderers to improve viewpoint robustness, but they typically depend on manually constructed 3D models. We introduce a semi-automated pipeline that generates physically printable and perspective-invariant adversarial patches using only a small set of 2D images. Our method integrates 3D reconstruction, neural rendering, adversarial patch optimization, and an object detection victim model into a unified workflow. We use 2D Gaussian Splatting for high fidelity mesh reconstruction and FlexPara for surface parameterization that produces texture maps suitable for patch editing. Together, these components form a fully differentiable pipeline in PyTorch3D that links texture modification to model outputs, enabling efficient optimization of patches that remain effective across many viewpoints. The complete process, from image capture to patch printing and physical evaluation, can be completed within a few hours. We demonstrate the effectiveness of the resulting patches through attacks on the YOLOv8 object detection model and discuss remaining challenges and opportunities for improving robustness and scalability.

Singhvi, Vivaan [ORNL] (ORCID:0009000586288221)↗

Risk-Sensitive Control of Pure Jump Process on Countable Space with Near Monotone Cost

In this article, we study risk-sensitive control problem with controlled continuous time pure jump process on a countable space as state dynamics. We prove multiplicative dynamic programming principle, elliptic and parabolic Harnack’s inequalities. Using the multiplicative dynamic programing principle and the Harnack’s inequalities, we prove the existence and a characterization of optimal risk-sensitive control under the near monotone condition.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

3D high-fidelity automated neutronics guided optimization of fusion blanket designs

The compact Fusion Pilot Plant (FPP) is defined in the recent National Academies of Sciences, Engineering, and Medicine report as the next step of fusion energy demonstration with a $50$ MWe peak net electricity production, $Q_e$ greater than $1$, and at least $3$ hours of continuous operation. This fusion pilot plant will be a test bed enabling materials, designs, and fuel management assessment, and it will represent an engineering challenge because of its high-fusion power and compact design targets. Previous reactor data is limited to experiments operating in different design space ranges. Therefore, design iterations and assessments should rely on high-fidelity first-principle theoretical and computational models. The high-fidelity integrated modeling of the plasma is a fundamental part of fusion energy research. However, the whole device modeling is often neglected, utilizing low-fidelity, system-level analysis. Recently, the need for high-fidelity multi-physics modeling was recognized, resulting in a selection of integrated tools. Further, autonomous design optimization requires a streamlined framework that perturbs the design point, reruns the analysis, and examines the outputs. However, high-fidelity analysis requires complex geometry specification that is difficult to perturb. This work presents the parametric CAD generation tool TRACER and a new neutronic workflow. TRACER allows the perturbation of the geometry representation, creating geometry files ready for further analysis. The streamlined neutronic workflow allows efficient and accurate calculations. The two new tools coupled together were used to perform a 3D high-fidelity multi-objective, multi-input optimization of an "ARC Class" compact tokamak design. The workflow was driven by an optimization driver for full automation.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Core Design Optimization of the Westinghouse Lead Fast Reactor

Westinghouse is pursuing an advanced Nuclear Power Plant design based on Lead Fast Reactor (LFR) technology for global commercialization. To achieve an optimal combination of key attributes, such as safety, sustainability, and economic competitiveness, Westinghouse and ANL partnered in developing and applying a formalized core design optimization strategy. An LFR analysis workflow was developed to automate a suite of reactor physics, fuels performance, safety, and economics simulations on a selected LFR concept. The workflow streamlines analysis of a wide range of LFR designs with different dimensions and fuel types to assess their viability and economic performance, significantly reducing human processing time and risks of processing errors. The LFR optimization exercise was defined, resulting in selection of the design constraints (geometric, neutronics, thermo-mechanical, safety, thermal-hydraulics, and economics) and performance metrics researched (minimization of both the fuels LCOE and the first core inventory cost). A total of 14 varied design parameters were considered, including assembly dimensions, coolant temperature, and enrichment distribution throughout the core. The LFR analysis workflow was connected to DAKOTA for sensitivity and optimization analyses. Due to the extremely large size of the potential LFR optimization solution space relative to the computing time required to characterize one LFR solution, a multi-stage optimization approach was proposed to breakdown the problem into several stages with more reasonable sizes. This optimization approach enabled finding various viable core solutions with different cost tradeoffs that were considered by Westinghouse and justify selection of a smaller core with multi-batch 2-year cycle length.

Stauff, Nicolas E.↗

Predicting the Evolution of Shallow Cumulus Clouds With a Lotka‐Volterra Like Model

Abstract In numerical weather prediction and climate models, boundary‐layer clouds are controlled by a wide range of subgrid‐scale processes. However, understanding the nature of these processes and their role in the evolution of the cloud size distribution as a whole has been elusive. To address this issue, we adopt a novel empirical framework from the field of population dynamics to model the evolution of cloud size statistics by using the shallow cumulus properties obtained from a large‐eddy simulation (LES). Our approach involves representing the cloud size distribution and the total cloud area using a revised Lotka‐Volterra model and ridge linear model, respectively. The physical interpretation of the total cloud area and coefficients obtained from the optimization of the models reveals three stages probably interpreted by dominant processes: the formation of new clouds, the growth of single clouds, and a steady state with organized transitions involving the growth and decay of multiple clouds. Furthermore, we showcase the potential of this framework to serve as a component of scale‐aware parameterizations of shallow‐convective clouds in atmospheric models.

54 ENVIRONMENTAL SCIENCES↗

On a Time-Optimal Control Problem Associated with the Heat Exchange Process

The initial-boundary problem for the heat conduction equation inside a bounded domain is considered. It is supposed that on the boundary of this domain the heat exchange according to Newton's law takes place. The control parameter is equal to the magnitude of output of hot or cold air and is defined on a given part of the boundary. An estimate of the minimal time for achieving the given average temperature is found.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

ENGINE COMBUSTION SYSTEM OPTIMIZATION USING CFD AND MACHINE LEARNING: A METHODOLOGICAL APPROACH

Gasoline compression ignition (GCI) engines are considered an attractive alternative to traditional spark-ignition and diesel engines. In this work, a Machine Learning-Grid Gradient Algorithm (ML-GGA) approach was developed to optimize the performance of internal combustion engines. Machine learning (ML) offers a pathway to transform complex physical processes that occur in a combustion engine into compact informational processes. The developed ML-GGA model was compared with a recently developed Machine learning Genetic Algorithm (ML-GA). Detailed investigations of optimization solver parameters and variables limits extension were performed in the present ML-GGA model to improve the accuracy and robustness of the optimization process. Detailed descriptions of the different procedures, optimization tools and criteria that must be followed for a successful output are provided here. The developed ML-GGA approach was used to optimize the operating conditions (case 1) and the piston bowl design (case 2) of a heavy-duty diesel engine running on a gasoline fuel with a Research Octane Number (RON) of 80. The ML-GGA approach yielded > 2% improvements in the merit function, compared to the optimum obtained from a thorough computational fluid dynamics (CFD) guided system optimization. The predictions from the ML-GGA approach were validated with engine CFD simulations. This study demonstrates the potential of ML-GGA to significantly reduce the time needed for optimization problems, without loss in accuracy compared to traditional approaches.

engine optimization↗

Device physics underlying silicon heterojunction and passivating‐contact solar cells: A topical review

Abstract The device physics of commercially dominant diffused‐junction silicon solar cells is well understood, allowing sophisticated optimization of this class of devices. Recently, so‐called passivating‐contact solar cell technologies have become prominent, with Kaneka setting the world's silicon solar cell efficiency record of 26.63% using silicon heterojunction contacts in an interdigitated configuration. Although passivating‐contact solar cells are remarkably efficient, their underlying device physics is not yet completely understood, not in the least because they are constructed from diverse materials that may introduce electronic barriers in the current flow. To bridge this gap in understanding, we explore the device physics of passivating contact silicon heterojunction (SHJ) solar cells. Here, we identify the key properties of heterojunctions that affect cell efficiency, analyze the dependence of key heterojunction properties on carrier transport under light and dark conditions, provide a self‐consistent multiprobe approach to extract heterojunction parameters using several characterization techniques (including dark J‐V, light J‐V, C‐V, admittance spectroscopy, and Suns‐V oc ), propose design guidelines to address bottlenecks in energy production in SHJ cells, and develop a process‐to‐module modeling framework to establish the module's performance limits. We expect that our proposed guidelines resulting from this multiscale and self‐consistent framework will improve the performance of future SHJ cells as well as other passivating contact‐based solar cells.

Chavali, Raghu V. K.↗

Characterization of zinc oxide thin film using atomic force microscopy and optimized X-ray reflectivity by genetic algorithm

For the nanostructured thin films, the study of surface physical parameters is often of crucial importance. For example, the transport properties of ZnO thin film are strongly affected by the geometry of this film. In this study sol—gel ZnO thin film was spin coated on glass substrate. The inverse problem of thin film characterization is formulated as a parameter identification problem in which a set of parameters corresponding to the thin film property can be found by minimizing fitness functions formulated using theoretical and experimental X-ray reflectivity. Atomic force microscopy (AFM) measurements used as an independent procedure for extracting structural parameters of the film and accurate interpretation of X-ray reflectivity data. After applying Fractal calculations on the AFM measured data the Genetic algorithm (GA) method was applied for extracting other important structural parameters. Optimization of specular X-ray reflectivity by GA method gives information along the depth such as thickness, roughness, and electron density profile of layers in the film.

36 MATERIALS SCIENCE↗

Data reduction through optimized scalar quantization for more compact neural networks

Raw data generation for several existing and planned large physics experiments now exceeds TB/s rates, generating untenable data sets in very little time. Those data often demonstrate high dimensionality while containing limited information. Meanwhile, Machine Learning algorithms are now becoming an essential part of data processing and data analysis. Those algorithms can be used offline for post processing and post data analysis, or they can be used online for real time processing providing ultra low latency experiment monitoring. Both use cases would benefit from data throughput reduction while preserving relevant information: one by reducing the offline storage requirements by several orders of magnitude and the other by allowing ultra fast online inferencing with low complexity Machine Learning models. Moreover, reducing the data source throughput also reduces material cost, power and data management requirements. In this work we demonstrate optimized nonuniform scalar quantization for data source reduction. This data reduction allows lower dimensional representations while preserving the relevant information of the data, thus enabling high accuracy Tiny Machine Learning classifier models for online fast inferences. We demonstrate this approach with an initial proof of concept targeting the CookieBox, an array of electron spectrometers used for angular streaking, that was developed for LCLS-II as an online beam diagnostic tool. We used the Lloyd-Max algorithm with the CookieBox dataset to design an optimized nonuniform scalar quantizer. Optimized quantization lets us reduce input data volume by 69% with no significant impact on inference accuracy. When we tolerate a 2% loss on inference accuracy, we achieved 81% of input data reduction. Finally, the change from a 7-bit to a 3-bit input data quantization reduces our neural network size by 38%.

97 MATHEMATICS AND COMPUTING↗

Perfect coupling of light to a periodic dielectric/metal/dielectric structure

Using the finite difference time domain method, it is demonstrated that perfect coupling can be achieved between normally incident light and a periodic dielectric/metal/dielectric structure. The structure serves as a diffraction grating that excites modes related to the long range surface plasmon and short range surface plasmon modes that propagate on continuous metallic films. By optimizing the structural dimensions, perfect coupling is achieved between the incident light and these modes. A high Q of 697 and an accompanying ultrasharp linewidth of 0.8 nm are predicted for a 10 nm silver film for optimal conditions.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

Performance of Multi-chaotic PSO on a shifted benchmark functions set

In this paper the performance of Multi-chaotic PSO algorithm is investigated using two shifted benchmark functions. The purpose of shifted benchmark functions is to simulate the time-variant real-world problems. The results of chaotic PSO are compared with canonical version of the algorithm. It is concluded that using the multi-chaotic approach can lead to better results in optimization of shifted functions.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Integrated modeling of electron cyclotron resonance ion sources and charge breeders with GEM, MCBC, and IonEx

A numerical toolset to help in understanding physical processes in the electron cyclotron resonance charge breeder (ECRCB) and further to help optimization and design of current and future machines is presented. The toolset consists of three modules (Monte Carlo charge breeding code, generalized electron cyclotron resonance ion source modeling, and ion extraction), each modeling different processes occurring in the ECRCB from beam injection to extraction. The toolset provides qualitative study, such as parameter studies, and scaling of the operation, and physical understanding in the ECRCB. The methodology and a sample integrated modeling are presented.

43 PARTICLE ACCELERATORS↗

Laser Acceleration of Quasi-Monoenergetic Protons via Radiation Pressure Driven Thin Foil

We present a theoretical and simulation study of laser acceleration of quasi-monoenergetic protons in a thin foil irradiated by high intensity laser light. The underlying physics of radiation pressure acceleration (RPA) is discussed, including the importance of optimal thickness and circularly polarized light for efficient acceleration of ions to quasi-monoenergetic beams. Preliminary two-dimensional simulation studies show that certain parameter regimes allow for stabilization of the Rayleigh-Taylor instability and possibility of acceleration of monoenergetic ions to an excess of 200 MeV, making them suitable for important applications such as medical cancer therapy and fast ignition.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗