Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “inversion methods”

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 757 records · Page 42
All sourcesNASA NTRSDOE OSTI
DOE OSTI2024

Effective interactions of the open bosonic string via field theory

Abstract We describe a method to extract an effective Lagrangian description for open bosonic strings, at zero transcendentality. The method relies on a particular formulation of its scattering amplitudes derived from color-kinematics duality. More precisely, starting from a (DF) 2 + YM quantum field theory, we integrate out all the massive degrees of freedom to generate an expansion in the inverse string tensionα′. We explicitly compute the Lagrangian terms through$$ \mathcal{O} $$ O (α′ 4 ), and target the sector of operators proportional toF 4 to all orders inα′.

Physics↗
DOE OSTI2022

SREMI: Super-resolution electromagnetic imaging with single-channel ground-penetrating radar

High-resolution near-surface imaging has important applications in civil engineering, infrastructure inspection, military threat detection, geological characterization, and lunar and planetary exploration. Zero-offset, singlechannel ground penetrating radar (GPR) imaging is an established technique for near-surface target imaging and sensing but often suffers from low spatial resolution and imaging artifacts, especially of deep structures. In response, we formulate the GPR imaging as a dual-sparsity optimization problem, and develop a super-resolution electromagnetic imaging method based on a fast iterative shrinkage-thresholding algorithm. We develop our GPR imaging method in the framework of electromagnetic exploding-reflectors simulation theory, therefore the imaging method is computationally efficient. In this work, we demonstrate through synthetic and field data examples that our method can produce sharper, more reliable images with fewer artifacts compared with single-pass reverse-time migration GPR method, thus leading to improved near-surface interpretation and object identification.

58 GEOSCIENCES↗
DOE OSTI2020

Generative adversarial networks (GAN) based efficient sampling of chemical composition space for inverse design of inorganic materials

A major challenge in materials design is how to efficiently search the vast chemical design space to find the materials with desired properties. One effective strategy is to develop sampling algorithms that can exploit both explicit chemical knowledge and implicit composition rules embodied in the large materials database. Here, we propose a generative machine learning model (MatGAN) based on a generative adversarial network (GAN) for efficient generation of new hypothetical inorganic materials. Trained with materials from the ICSD database, our GAN model can generate hypothetical materials not existing in the training dataset, reaching a novelty of 92.53% when generating 2 million samples. The percentage of chemically valid (charge-neutral and electronegativity-balanced) samples out of all generated ones reaches 84.5% when generated by our GAN trained with such samples screened from ICSD, even though no such chemical rules are explicitly enforced in our GAN model, indicating its capability to learn implicit chemical composition rules to form compounds. Our algorithm is expected to be used to greatly expand the range of the design space for inverse design and large-scale computational screening of inorganic materials.

36 MATERIALS SCIENCE↗
DOE OSTI2023

Design of refractory multi-principal-element alloys for high-temperature applications

Abstract Refractory multi-principal-element alloys (RMPEAs) exhibit high specific strength at elevated temperatures ( T ). However, current RMPEAs lack a balance of room-temperature (RT) ductility, high- T strength, and high- T creep resistance. Using density-functional theory methods, we scanned composition space using four criteria: (1) formation energies for operational stability: $$-150\le {E}_{{\rm {f}}}$$ − 150 ≤ E f ≤ +70 meV per atom; (2) higher strength found via interstitial electron density with Young’s moduli E > 250 GPa; (3) inverse Pugh ratio for ductility: G / B < 0.57; and (4) high melting points: T m > 2500 °C. Using rapid bulk alloy synthesis and characterization, we validated theory and down-selected promising alloy compositions and discovered Mo 72.3 W 12.8 Ta 10.0 Ti 2.5 Zr 2.5 having well-balanced RT and high- T mechanical properties. This alloy has comparable high- T compressive strength to well-known MoNbTaW but is more ductile and more creep resistant. It is also superior to a commercial Mo-based refractory alloy and a nickel-based superalloy (Haynes-282) with improved high- T tensile strength and creep resistance.

36 MATERIALS SCIENCE↗
DOE OSTI2023

Machine learning-based real-time kinetic profile reconstruction in DIII-D

Abstract Kinetic equilibrium reconstruction plays a vital role in the physical analysis of plasma stability and control in fusion tokamaks. However, the traditional approach is subjective and prone to human biases. To address this, the consistent automatic kinetic equilibrium reconstruction (CAKE) method was introduced, providing objective results. Nonetheless, its offline nature limits its application in real-time plasma control systems (PCSs). To address this limitation, we present RTCAKENN, a machine learning model that approximates 7 CAKE-level output profiles, namely pressure, inverse q , toroidal current density, electron temperature and density, carbon ion impurity temperature and rotation profiles, using real-time available inputs. The deep neural network consists of an encoder layer, where the scalars and interdependent inputs such as plasma boundary coordinates and motional Stark effect data are encoded using multi-layer perceptrons (MLPs), while profile inputs are encoded by 1D convolutional layers. The encoded data is passed through a MLP for latent feature extraction, before being decoded in the decoding layers, which consist of upsampling and convolutional layers. RTCAKENN has been implemented in the DIII-D PCS and our model achieves accuracy comparable to CAKE and surpasses existing real-time alternatives. Through clever dropout training, RTCAKENN exhibits robustness and can operate even in the absence of Thomson scattering data or charge exchange recombination data. It executes in under 8 ms in the real-time environment, enabling future application in real-time control and analysis.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2026

Shape evolution in neutron-rich odd-even 105–109 Nb isotopes

Background: Neutron-rich nuclei around 𝑍 ≈ 40 are well known for exhibiting multiple shape transitions. Here, this region shows one of the sharpest shape transitions in the nuclear chart, evolving from a spherical vibrator at 𝑁 = 58 to a strongly deformed prolate shape at 𝑁 = 60. The largest deformations are observed for 38 Sr and 40 Zr . This abrupt shape transition disappears at 𝑍 = 36 and below, where a shape transition from spherical to oblate nuclei is predicted. On the other hand, for 𝑍 ≥ 42 and 𝑁 ≥ 60, the shape is known to evolve from axial to triaxial. While the even-𝑍 nuclei in this region have already been extensively studied, new insights can be gained from the studies of odd-𝑍 isotopes for a better understanding of the underlying mechanisms driving these phenomena. Purpose: The 41 Nb nuclei lie at the boundary between axially deformed Zr and triaxially deformed Mo nuclei. This work investigates the nuclear structure of very neutron-rich Nb nuclei up to 𝑁 = 68. The goal is to understand how the nuclear shape evolves as a function of isospin in this isotopic chain and provide new insights into the emergence of triaxial deformation. Methods: The structure of the neutron-rich Nb isotopes was investigated using state-of-the-art high-resolution 𝛾-ray spectroscopy of fission fragments produced via two different fission reactions. The use of 9 Be ⁢( 238 U, 𝑓) inverse kinematics, with a detection system comprising AGATA, EXOGAM, and VAMOS++, enabled the measurement of prompt and delayed 𝛾 rays from isotopically identified fission fragments, and 𝛾−𝛾−𝛾−𝛾 highfold data were obtained from a spontaneous fission source of 252 Cf using the Gammasphere array. Results: The level scheme of 105 Nb has been significantly extended, with the addition of two negative-parity bands observed for the first time. A new level scheme is proposed for 107 Nb, which is not in agreement with an earlier measurement, and new levels and transitions have been added to the very neutron-rich 109 Nb. The degree of triaxiality of the new bands is discussed on the basis of signature splitting analysis. The recently reported level scheme of 99 Nb has been revised. Conclusions: This systematic study on the Nb isotopic chain, compared to Zr and Mo, indicates that while the ground-state band exhibits a triaxial deformation, attributed to a proton hole coupled to a triaxially deformed Mo core, the negative-parity bands, based on isomeric bandheads, display an axially symmetric deformed structure, similar to that observed in the Zr isotopes, revealing the existence of a shape coexistence in the neutron-rich Nb nuclei.

Abushawish, M. [Université Claude-Bernard Lyon 1 (↗
DOE OSTI2026

Regularizing INR with Diffusion Prior for Self-Supervised 3D Reconstruction OF Neutron Computed Tomography Data

Recently, generative diffusion priors have made huge strides as inverse problem solvers, including the ability to be adapted for inference on out-of-distribution data. Concurrently, implicit neural representations (INRs) have emerged as fast and lightweight inverse imaging solvers that are amenable to hybrid approaches that combine learned priors with traditional inverse problem formulations. In this paper, we present a diffusive computed tomography (CT) inversion framework for regularizing INRs called Diffusive INR (DINR), designed to enable high-quality reconstruction from sparse-view neutron CT. Pretrained purely on synthetic data, DINR is evaluated on simulated and experimentally obtained observations of concrete microstructures, where traditional reconstruction methods suffer substantial degradation when the number of views is reduced. Our approach delivers superior performance, reduces reconstruction artifacts, and achieves gains in PSNR and SSIM, enabling accurate micro-structural characterization even under extreme data limitations compared to state-of-the-art sparse-view reconstruction techniques.

Hossain, Maliha [ORNL]↗
DOE OSTI2020

Towards real-time monitoring: data assimilated time-lapse full waveform inversion for seismic velocity and uncertainty estimation

SUMMARY Rapid development of time-lapse seismic monitoring instrumentations has made it possible to collect dense time-lapse data for tomographically retrieving time-lapse (even continuous) images of subsurface changes. While traditional time-lapse full waveform inversion (TLFWI) algorithms are designed for sparse time-lapse surveys, they lack of effective temporal constraint on time-lapse data, and, more importantly, lack of the uncertainty estimation of the TLFWI results that is critical for further interpretation. Here, we propose a new data assimilation TLFWI method, using hierarchical matrix powered extended Kalman filter (HiEKF) to quantify the image uncertainty. Compared to existing Kalman filter algorithms, HiEKF allows to store and update a data-sparse representation of the cross-covariance matrices and propagate model errors without expensive operations involving covariance matrices. Hence, HiEKF is computationally efficient and applicable to 3-D TLFWI problems. Then, we reformulate TLFWI in the framework of HiEKF (termed hereafter as TLFWI-HiEKF) to predict time-lapse images of subsurface spatiotemporal velocity changes and simultaneously quantify the uncertainty of the inverted velocity changes over time. We demonstrate the validity and applicability of TLFWI–HiEKF with two realistic CO2 monitoring models derived from Frio-II and Cranfield CO2 injection sites, respectively. In both 2-D and 3-D examples, the inverted high-resolution time-lapse velocity results clearly reveal a continuous velocity reduction due to the injection of CO2. Moreover, the accuracy of the model is increasing over time by assimilating more time-lapse data while the standard deviation is decreasing over lapsed time. We expect TLFWI-HiEKF to be equipped with real-time seismic monitoring systems for continuously imaging the distribution of subsurface gas and fluids in the future large-scale CO2 sequestration experiments and reservoir management.

58 GEOSCIENCES↗
DOE OSTI2024

Mapping Permafrost Variability and Degradation Using Seismic Surface Waves, Electrical Resistivity, and Temperature Sensing: A Case Study in Arctic Alaska

Subsurface processes significantly influence surface dynamics in permafrost regions, necessitating utilizing diverse geophysical methods to reliably constrain permafrost characteristics. This research uses multiple geophysical techniques to explore the spatial variability of permafrost in undisturbed tundra and its degradation in disturbed tundra in Utqiagvik, Alaska. Here, we integrate multiple quantitative techniques, including multichannel analysis of surface waves (MASW), electrical resistivity tomography (ERT), and ground temperature sensing, to study heterogeneity in permafrost’s geophysical characteristics. MASW results reveal active layer shear wave velocities (V s ) between 240 and 370 m/s, and permafrost V s between 450 and 1,700 m/s, typically showing a low-high-low velocity pattern. Additionally, we find an inverse relationship between in situ V s and ground temperature measurements. The V s profiles along with electrical resistivity profiles reveal cryostructures such as cryopeg and ice-rich zones in the permafrost layer. The integrated results of MASW and ERT provide valuable information for characterizing permafrost heterogeneity and cryostructure. Corroboration of these geophysical observations with permafrost core samples’ stratigraphies and salinity measurements further validates these findings. This combination of geophysical and temperature sensing methods along with permafrost core sampling confirms a robust approach for assessing permafrost’s spatial variability in coastal environments. Our results also indicate that civil infrastructure systems such as gravel roads and pile foundations affect permafrost by thickening the active layer, lowering the V s , and reducing heterogeneity. We show how the resulting V s profiles can be used to estimate key parameters for designing buildings in permafrost regions and maintaining existing infrastructure in polar regions.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2022

Highly Selective Methane to Methanol Conversion on Inverse SnO 2 /Cu 2 O/Cu(111) Catalysts: Unique Properties of SnO 2 Nanostructures and the Inhibition of the Direct Oxidative Combustion of Methane

Direct methane to methanol (CH 4 → CH 3 OH) conversion in heterogeneous catalysis has been a long-standing challenge due to the difficulties in equalizing the activation of methane and protection of the methanol product at the same reaction conditions. Here, we report an inverse catalyst, consisting of small structures of SnO 2 (0.5-1 nm in size) dispersed on Cu 2 O/Cu(111), for highly selective CH 3 OH production from CH 4 . This system was investigated by combining theoretical [density functional theory calculations (DFT), kinetic Monte Carlo simulations (KMC)] and experimental methods [scanning tunneling microscopy (STM), ambient-pressure X-ray photoelectron spectroscopy (AP-XPS)]. The DFT and AP-XPS studies showed that on SnO 2 /Cu 2 O/Cu(111) the conversion of CH 4 by oxygen (O 2 ) preferred complete combustion to carbon dioxide (CO 2 ). The addition of water (H 2 O) enhanced the production of CH 3 OH to nearly 100% selectivity in KMC simulations. This trend was consistent with results of AP-XPS. The presence of water in the reaction environment rendered an extremely high amount of methoxy species (*CH 3 O), a precursor for CH 3 OH production. Further, the high CH 3 OH selectivity of SnO 2 /Cu 2 O/Cu(111) reflected the unique atomic and electronic structure of the supported SnO 2 nanoparticles. As a result, the O 2 adsorption and dissociation, and thus the full combustion of CH 4 to CO 2 , was completely suppressed; while the H 2 O dissociative adsorption was still feasible, providing active hydroxyl species for a truly selective CH 4 to CH 3 OH conversion.

03 NATURAL GAS↗
DOE OSTI2020

Low synchronization Gram–Schmidt and generalized minimal residual algorithms

The Gram–Schmidt process uses orthogonal projection to construct the A = QR factorization of a matrix. When Q has linearly independent columns, the operator P = I - Q(QTQ)-1QT defines an orthogonal projection onto Q⊥. In finite precision, Q loses orthogonality as the factorization progresses. A family of approximate projections is derived with the form P = I - QTQT, with correction matrix T. When T = (QTQ)-1, and T is triangular, it is postulated that the best achievable orthogonality is $\mathcal{O}(ε)\mathcal{K}(A)$. We present new variants of modified (MGS) and classical Gram–Schmidt algorithms that require one global reduction step. An interesting form of the projector leads to a compact WY representation for MGS. In particular, the inverse compact WY MGS algorithm is equivalent to a lower triangular solve. Our main contribution is to introduce a backward normalization lag into the compact WY representation, resulting in a $\mathcal{O}(ε)\mathcal{K}[r_0, AV_m])$ stable Generalized Minimal Residual Method (GMRES) algorithm that requires only one global reduce per iteration. Finally, further improvements in performance are achieved by accelerating GMRES on GPUs.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2026

Beyond interpolation: Physics-inspired gating transformers for extrapolating irradiation conditions to novel nuclear fuels

The qualification of advanced nuclear fuels relies on irradiation experiments in test reactors that emulate commercial conditions. Designing these tests requires accurate prediction of key irradiation quantities, particularly heat generation rate and burnup, yet obtaining them typically involves computationally expensive multi-step simulation workflows. We propose a physics-inspired gating transformer (PIGT) that integrates an inverse-square, distance-based attenuation into the encoder representation to bias attention toward physically relevant spatial relationships while retaining data-driven flexibility. Using MiniFuel irradiation data from the High Flux Isotope Reactor at Oak Ridge National Laboratory, we benchmark against ensemble methods, feedforward and recurrent networks, convolutional models, and standard transformers. While baseline models perform well under interpolation, they exhibit a pronounced generalization gap when evaluated on fuels not included in the training set. The proposed model consistently improves extrapolative accuracy and stability, yielding the strongest performance on unseen fuel configurations. These results indicate that a lightweight physics structure embedded within attention mechanisms can substantially improve robustness, enabling more reliable surrogate predictions to accelerate the design of nuclear fuel irradiation experiments.

Fuel qualification↗
DOE OSTI2026

Laplace Transform–Based Quantum Eigenvalue Transformation via Linear Combination of Hamiltonian Simulation

Eigenvalue transformations, which include solving time-dependent differential equations as a special case, have a wide range of applications in scientific and engineering computation. While quantum algorithms for singular value transformations are well studied, eigenvalue transformations are distinct, especially for nonnormal matrices. Here, we propose an efficient quantum algorithm for performing a class of eigenvalue transformations that can be expressed as a certain type of matrix Laplace transformation. This allows us to significantly extend the recently developed linear combination of Hamiltonian simulation method [D. An, J.-P. Liu, and L. Lin, Phys. Rev. Lett., 131 (2023), 150603; D. An, A. M. Childs, and L. Lin, Commun. Math. Phys. 407, 19 (2026)] to represent a wider class of eigenvalue transformations, such as powers of the matrix inverse, 𝐴 −𝑘 , and the exponential of the matrix inverse, 𝑒 −𝐴 −1 . The latter can be interpreted as the solution of a mass-matrix differential equation of the form form 𝐴⁢𝑢′⁡⁡(𝑡) =−𝑢⁡(𝑡). We demonstrate that our eigenvalue transformation approach can solve this problem without explicitly inverting 𝐴, thereby reducing the computational complexity.

Laplace transform↗
DOE OSTI2024

ML-based Dimension Reduction Strategies

Deep learning (DL)--based surrogate models have achieved success in various applications in carbon capture and storage (CCS). However, the model training on high-dimensional spaces is computationally expensive and impractical for large-scale and complex geological models, because the models usually contain hundreds of thousands to millions of grid cells, each with a set of parameters. Furthermore, the high cost of generating training data with sufficient variation is another limitation of model training on high-dimensional spaces, which may result in overfitting and reduce the model efficiency and prediction performance. We proposed the workflow incorporating dimension reduction methods and deep learning models, which aim to extract the latent variables of input parameters and output state variables, and then build the mapping function at the latent spaces. The proposed workflow can significantly reduce the computational complexity in solving both forward and inverse problems compared to models trained on high-dimensional spaces. Dimensionality reduction models showed great potential in workflows for fast reservoir simulation, history matching, prior model generation, visualization, and more, ultimately enhancing DL model performance in related SMART Work Packages.

Hosseini, Seyyed↗
DOE OSTI2021

Reconstructing Masses of Merging Neutron Stars from Stellar r -process Abundance Signatures

We report that neutron star mergers (NSMs) are promising astrophysical sites for the rapid neutron-capture ("r") process, but can their integrated yields explain the majority of heavy-element material in the Galaxy? One method to address this question implements a forward approach that propagates NSM rates and yields along with stellar formation rates and compares those results with observed chemical abundances of r-process-rich, metal-poor stars. In this work, we take the inverse approach by utilizing r-process-element abundance ratios of metal-poor stars as input to reconstruct the properties -especially the masses - of their neutron star (NS) binary progenitors. This novel analysis provides an independent avenue for studying the population of the original NS binary systems that merged and produced the r-process material now incorporated in Galactic metal-poor halo stars. We use ratios of elements typically associated with the limited-r-process and the actinide region to those in the lanthanide region (i.e., Zr/Dy and Th/Dy) to probe the NS masses of the progenitor merger. We find that NSMs can account for all r-process material in metal-poor stars that display r-process signatures, while simultaneously reproducing the present-day distribution of double-NS systems. Notably, with our model assumptions and the studied stellar sample, we postulate that the most r-process enhanced stars (the r–II stars) on their own would require progenitor NSMs of asymmetric systems that are distinctly different from present ones in the Galaxy. We also explore variations to the model and find that the predicted degree of asymmetry is most sensitive to the electron fraction of the remnant disk wind.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2025

Inverse Trans Influence and Uranium-Arene σ-Bonding Drive Molecular Geometry: Ligand Modification from Hard to Soft Flips the Oxide from Axial to Equatorial

A rare example of an equatorially bound terminal uranium(V) oxo complex in a chelating sulfur-based ligand environment, namely [(mes( Me,Ad ArS) 3 )U V (O eq )(THF)] (2), is presented. Octahedrally coordinated 2 is obtained by reaction of the mesitylene-anchored tris-thiophenolate-coordinated uranium(III) complex [U III ((SAr Ad,Me ) 3 mes)] (1) with the oxygen-atom transfer reagent N 2 O. The observed, equatorially bound oxo ligand in 2 is in stark contrast to its known tris-aryloxide analog, [(mes( Me,Ad ArO) 3 )U V (O ax )(THF)] (A), where the oxo ligand occupies the typically observed axial coordination site. Complexes 1 and 2 are characterized by single-crystal X-ray diffraction analyses and spectroscopic and magnetochemical methods, including 1 H NMR, UV/vis/NIR electronic absorption, as well as EPR spectroscopy and SQUID magnetometry, thus confirming the C S symmetry and the pentavalent oxidation state of 2. Encompassing quantum chemical calculations (DFT and CASPT2) on 2 and its tris-phenolate analog A, support and rationalize the structural and electronic differences. The molecular orbital pictures show that a stabilizing σ-bonding interaction arising from the U–O eq inverse trans influence (ITI) is present in 2 but missing in A. In 2, the sulfur 3p orbitals are closer in energy to the uranium 5f manifold than the arene π-system, leading to an ITI, while U–arene σ- or δ-bonding is not observed. Although the arene orbitals remain separated from the uranium 5f orbitals in A, the absence of an ITI allows the arene a 2u orbital to engage in a σ-type interaction with the metal. Thus, incorporating a tris-thiophenolate to an arene anchor introduces a new design concept in molecular f-element chemistry. This approach stabilizes an equatorially bound U(V) oxo center, contrasting with its tris-phenolate counterpart, where oxo coordination is axial. The observed geometric divergence, driven by competing ITI and U–arene interactions, not only tunes electronic structure but also leads to differentiated reactivity: only the phenolate analogs activate H 2 O, while the thiolates do not.

Hydrocarbons↗
DOE OSTI2023

Conditional Karhunen–Loève regression model with Basis Adaptation for high-dimensional problems: Uncertainty quantification and inverse modeling

Here, we propose a methodology for improving the accuracy of surrogate models of the observable response of physical systems as a function of the systems’ spatially heterogeneous parameter fields, with applications to uncertainty quantification and parameter estimation in high-dimensional problems. Practitioners often formulate finite-dimensional representations of spatially heterogeneous parameter fields using truncated unconditional Karhunen–Loève expansions (KLEs) for a certain choice of unconditional covariance kernel and construct surrogate models of the observable response with respect to the KLE coefficients. When direct measurements of the parameter fields are available, we propose improving the accuracy of these surrogate models by representing the parameter fields via conditional Karhunen-Loève expansions (CKLEs). CKLEs are constructed by conditioning the covariance kernel of the unconditional expansion on the direct measurements of the parameter field via Gaussian process regression, and then truncating the corresponding KLE. We apply the proposed methodology to constructing surrogate models via the Basis Adaptation (BA) method of the stationary hydraulic head response, measured at spatially discrete observation locations, of a groundwater flow model of the Hanford Site, as a function of the 1000-dimensional representation of the model’s log-transmissivity field. We find that BA surrogate models of the hydraulic head based on CKLEs are more accurate than BA surrogate models based on unconditional expansions for forward uncertainty quantification tasks. Furthermore, we find that inverse estimates of the hydraulic transmissivity field computed using CKLE-based BA surrogate models are more accurate than those computed using unconditional BA surrogate models.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2021

INTEGRATE - Inverse Network Transformations for Efficient Generation of Robust Airfoil and Turbine Enhancements

The INTEGRATE (Inverse Network Transformations for Efficient Generation of Robust Airfoil and Turbine Enhancements) project is developing a new inverse-design capability for the aerodynamic design of wind turbine rotors using invertible neural networks. This AI-based design technology can capture complex non-linear aerodynamic effects while being 100 times faster than design approaches based on computational fluid dynamics. This project enables innovation in wind turbine design by accelerating time to market through higher-accuracy early design iterations to reduce the levelized cost of energy. INVERTIBLE NEURAL NETWORKS Researchers are leveraging a specialized invertible neural network (INN) architecture along with the novel dimension-reduction methods and airfoil/blade shape representations developed by collaborators at the National Institute of Standards and Technology (NIST) learns complex relationships between airfoil or blade shapes and their associated aerodynamic and structural properties. This INN architecture will accelerate designs by providing a cost-effective alternative to current industrial aerodynamic design processes, including: - Blade element momentum (BEM) theory models: limited effectiveness for design of offshore rotors with large, flexible blades where nonlinear aerodynamic effects dominate - Direct design using computational fluid dynamics (CFD): cost-prohibitive - Inverse-design models based on deep neural networks (DNNs): attractive alternative to CFD for 2D design problems, but quickly overwhelmed by the increased number of design variables in 3D problems AUTOMATED COMPUTATIONAL FLUID DYNAMICS FOR TRAINING DATA GENERATION - MERCURY FRAMEWORK The INN is trained on data obtained using the University of Marylands (UMD) Mercury Framework, which has with robust automated mesh generation capabilities and advanced turbulence and transition models validated for wind energy applications. Mercury is a multi-mesh paradigm, heterogeneous CPU-GPU framework. The framework incorporates three flow solvers at UMD, 1) OverTURNS, a structured solver on CPUs, 2) HAMSTR, a line based unstructured solver on CPUs, and 3) GARFIELD, a structured solver on GPUs. The framework is based on Python, that is often used to wrap C or Fortran codes for interoperability with other solvers. Communication between multiple solvers is accomplished with a Topology Independent Overset Grid Assembler (TIOGA). NOVEL AIRFOIL SHAPE REPRESENTATIONS USING GRASSMAN SPACES We developed a novel representation of shapes which decouples affine-style deformations from a rich set of data-driven deformations over a submanifold of the Grassmannian. The Grassmannian representation as an analytic generative model, informed by a database of physically relevant airfoils, offers (i) a rich set of novel 2D airfoil deformations not previously captured in the data , (ii) improved low-dimensional parameter domain for inferential statistics informing design/manufacturing, and (iii) consistent 3D blade representation and perturbation over a sequence of nominal shapes. TECHNOLOGY TRANSFER DEMONSTRATION - COUPLING WITH NREL WISDEM Researchers have integrated the inverse-design tool for 2D airfoils (INN-Airfoil) into WISDEM (Wind Plant Integrated Systems Design and Engineering Model), a multidisciplinary design and optimization framework for assessing the cost of energy, as part of tech-transfer demonstration. The integration of INN-Airfoil into WISDEM allows for the design of airfoils along with the blades that meet the dynamic design constraints on cost of energy, annual energy production, and the capital costs. Through preliminary studies, researchers have shown that the coupled INN-Airfoil + WISDEM approach reduces the cost of energy by around 1% compared to the conventional design approach. This page will serve as a place to easily access all the publications from this work and the repositories for the software developed and released through this pr...

aerodynamics↗