Search NASA⌕ Search

SEARCH · Search NASA

Results for “physics-based regularization”

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 19 records

A kinetic-based regularization method for data science applications

We propose a physics-based regularization technique for function learning, inspired by statistical mechanics. By drawing an analogy between optimizing the parameters of an interpolator and minimizing the energy of a system, we introduce corrections that impose constraints on the lower-order moments of the data distribution. This minimizes the discrepancy between the discrete and continuum representations of the data, in turn allowing to access more favorable energy landscapes, thus improving the accuracy of the interpolator. Our approach improves performance in both interpolation and regression tasks, even in high-dimensional spaces. Unlike traditional methods, it does not require empirical parameter tuning, making it particularly effective for handling noisy data. We also show that thanks to its local nature, the method offers computational and memory efficiency advantages over Radial Basis Function interpolators, especially for large datasets.

97 MATHEMATICS AND COMPUTING↗

NuGraph2 with context-aware inputs: physics-inspired improvements in semantic segmentation

Graph neural networks have recently shown strong promise for event reconstruction tasks in Liquid Argon Time Projection Chambers, yet their performance remains limited for underrepresented classes of particles, such as Michel electrons. In this work, we investigate physics-informed strategies to improve semantic segmentation within the NuGraph2 architecture. We explore three complementary approaches: (i) enriching the input representation with context-aware features derived from detector geometry and track continuity, (ii) introducing auxiliary decoders to capture class-level correlations, and (iii) incorporating energy-based regularization terms motivated by Michel electron energy distributions. Experiments on MicroBooNE public datasets show that physics-inspired feature augmentation yields the largest gains, particularly boosting Michel electron precision and recall by disentangling overlapping latent space regions. In contrast, auxiliary decoders and energy-regularization terms provided limited improvements, partly due to the hit-level nature of NuGraph2, which lacks explicit particle- or event-level representations. Our findings highlight that embedding physics context directly into node-level inputs is more effective than imposing task-specific auxiliary losses, and suggest that future hierarchical architectures such as NuGraph3, with explicit particle- and event-level reasoning, will provide a more natural setting for advanced decoders and physics-based regularization. The code for this work is publicly available on Github at https://github.com/vitorgrizzi/nugraph_phys/tree/main_phys.

Other Experiments↗

Constructing Neural Network Based Models for Simulating Dynamical Systems

Dynamical systems see widespread use in natural sciences like physics, biology, and chemistry, as well as engineering disciplines such as circuit analysis, computational fluid dynamics, and control. For simple systems, the differential equations governing the dynamics can be derived by applying fundamental physical laws. However, for more complex systems, this approach becomes exceedingly difficult. Data-driven modeling is an alternative paradigm that seeks to learn an approximation of the dynamics of a system using observations of the true system. In recent years, there has been an increased interest in applying data-driven modeling techniques to solve a wide range of problems in physics and engineering. Here this article provides a survey of the different ways to construct models of dynamical systems using neural networks. In addition to the basic overview, we review the related literature and outline the most significant challenges from numerical simulations that this modeling paradigm must overcome. Based on the reviewed literature and identified challenges, we provide a discussion on promising research areas.

97 MATHEMATICS AND COMPUTING↗

DENSECL: Haze Mitigation Using Dense Blocks and Contrastive Loss Regularization

Haze, which occurs as a result of the scattering of light in the atmosphere by small particles, diminishes the visibility of scene objects, inflicting important image applications such as object detection. To address the problem, this paper introduces a new physics-based end-to-end deep learning approach to haze mitigation in outdoor scenes, including those in airborne images. The proposed model named DenseCL is designed with dense blocks and adopts a contrastive loss function as an additional regularization. The model also maintains the cycle consistency by remapping the dehazed outputs into a hazy image using the physics-based light scattering function. DenseCL has been trained with publicly available outdoor images and demonstrates outstanding performance on outdoor, indoor, and remotely sensed nonhomogeneous haze satellite images.

Mitra, Somosmita↗

Deep-learning-based workflow for boundary and small target segmentation in digital rock images using UNet++ and IK-EBM

We report three-dimensional (3D) X-ray micro-computed tomography (μCT) has been widely used in petroleum engineering because it can provide detailed pore structural information for a reservoir rock, which can be imported into a pore-scale numerical model to simulate the transport and distribution of multiple fluids in the pore space. The partial volume blurring (PVB) problem is a major challenge in segmenting raw μCT images of rock samples, which impacts boundaries and small targets near the resolution limit. We developed a deep-learning (DL)-based workflow for accurate and fast partial volume segmentation. The DL model's performance depends primarily on the training data quality and model architecture. This study employed the entropy-based-masking indicator kriging (IK-EBM) to segment 3D Berea sandstone images as training datasets. The comparison between IK-EBM and manual segmentation using a 3D synthetic sphere pack, which had a known ground truth, showed that IK-EBM had higher accuracy on partial volume segmentation. We then trained and tested the UNet++ model, a state-of-the-art supervised encoder-decoder model, for binary (i.e., void and solid) and four-class segmentation. We compared the UNet++ with the commonly used U-Net and wide U-Net models and showed that the UNet++ had the best performance in terms of pixel-wise and physics-based evaluation metrics. Specifically, boundary-scaled accuracy demonstrated that the UNet++ architecture outperformed the regular U-Net architecture in the segmentation of pixels near boundaries and small targets, which were subjected to the PVB effect. Feature map visualization illustrated that the UNet++ bridged the semantic gaps between the feature maps extracted at different depths of the network, thereby enabling faster convergence and more accurate extraction of fine-scale features. The developed workflow significantly enhances the performance of supervised encoder-decoder models in partial volume segmentation, which has extensive applications in fundamental studies of subsurface energy, water, and environmental systems.

02 PETROLEUM↗

The National Solar Radiation Database (NSRDB) Fiscal Years 2019-2021(Final Report)

The National Solar Radiation Database (NSRDB) is the leading public source of high-resolution solar resource data in the United States, with more than 166,000 users annually. This database represents the state of the art in satellite-based estimation of solar resource information and uses a unique physics-based modeling approach that enables improvements in accuracy with the deployment of the next-generation geostationary satellites. Making the highest quality, state-of-the-art, regularly updated data sets available on a timely basis for users reduces costs of solar deployment by providing accurate information for siting studies and system output prediction, and thereby reduces levelized cost of energy. Also, high-resolution information from the NSRDB enables moving beyond levelized cost of energy when valuing the impact of renewables on the grid. Additionally, the NSRDB enables the integration of large amounts of solar on the grid by providing critical information about solar availability and variability that is used to enhance grid reliability and power quality.

14 SOLAR ENERGY↗

The National Solar Radiation Database Final Report: Fiscal Years 2022-2024

The National Solar Radiation Database (NSRDB) is the leading public source of high-resolution solar resource data in the United States, with more than 400,000 users annually. This database represents the state of the art in the satellite-based estimation of solar resource information and uses a unique physics-based modeling approach that enables improvements in accuracy with the deployment of the next-generation geostationary satellites. Making the highest-quality, state-of-the-art, regularly updated datasets available on a timely basis for users reduces costs of solar deployment by providing accurate information for siting studies and system output prediction and thereby reduces project financing costs and risks.

14 SOLAR ENERGY↗

A gradient-based deep neural network model for simulating multiphase flow in porous media

We report simulation of multiphase flow in porous media is crucial for the effective management of subsurface energy and environment-related activities. The numerical simulators used for modeling such processes rely on spatial and temporal discretization of the governing mass and energy balance partial-differential equations (PDEs) into algebraic systems via finite-difference/volume/element methods. These simulators usually require dedicated software development and maintenance, and suffer low efficiency from a runtime and memory standpoint for problems with multi-scale heterogeneity, coupled-physics processes or fluids with complex phase behavior. Therefore, developing cost-effective, data-driven models can become a practical choice, and in this work, we choose deep learning approaches as they can handle high dimensional data and accurately predict state variables with strong nonlinearity. In this paper, we describe a gradient-based deep neural network (GDNN) constrained by the physics related to multiphase flow in porous media. We tackle the nonlinearity of flow in porous media induced by rock heterogeneity, fluid properties, and fluid-rock interactions by decomposing the nonlinear PDEs into a dictionary of elementary differential operators. We use a combination of operators to handle rock spatial heterogeneity and fluid flow by advection. Since the augmented differential operators are inherently related to the physics of fluid flow, we treat them as first principles prior knowledge to regularize the GDNN training. We use the example of pressure management at geologic CO 2 storage sites, where CO 2 is injected in saline aquifers and brine is produced, and apply GDNN to construct a predictive model that is trained with physics-based simulation data and emulates the physics process. We demonstrate that GDNN can effectively predict the nonlinear patterns of subsurface responses, including the temporal and spatial evolution of the pressure and saturation plumes. We also successfully extend the GDNN to convolutional neural network (CNN), namely gradient-based CNN (GCNN), and validate its capability to improve the prediction accuracy. GDNN has great potential to tackle challenging problems that are governed by highly nonlinear physics and enable the development of data-driven models with higher fidelity.

42 ENGINEERING↗

PINN surrogate of Li-ion battery models for parameter inference, Part II: Regularization and application of the pseudo-2D model

Bayesian parameter inference is useful to improve Li-ion battery diagnostics and can help formulate battery aging models. However, it is computationally intensive and cannot be easily repeated for multiple cycles, multiple operating conditions, or multiple replicate cells. To reduce the computational cost of Bayesian calibration, numerical solvers for physics-based models can be replaced with faster surrogates. A physics-informed neural network (PINN) is developed as a surrogate for the pseudo-2D (P2D) battery model calibration. For the P2D surrogate, additional training regularization was needed as compared to the PINN single-particle model (SPM) developed in Part I. Both the PINN SPM and P2D surrogate models are exercised for parameter inference and compared to data obtained from a direct numerical solution of the governing equations. A parameter inference study highlights the ability to use these PINNs to calibrate scaling parameters for the cathode Li diffusion and the anode exchange current density. By realizing computational speed-ups of ~2250x for the P2D model, as compared to using standard integrating methods, the PINN surrogates enable rapid state-of-health diagnostics. Finally, in the low-data availability scenario, the testing error was estimated to ~2 mV for the SPM surrogate and ~10 mV for the P2D surrogate which could be mitigated with additional data.

25 ENERGY STORAGE↗

Inter-well connectivity detection in CO 2 WAG projects using statistical recurrent unit models

Routine well-wise injection and production measurements contain significant information on subsurface structure and properties. Data-driven technology that interprets surface data into subsurface structure or properties can assist operators in making informed decisions by providing a better understanding of field assets. Our machine-learning framework is built on the statistical recurrent unit (SRU) model and interprets well-based injection/production data into inter-well connectivity without relying on a geologic model. We test it on synthetic and field-scale CO 2 EOR projects utilizing the water-alternating-gas (WAG) process. SRU is a special type of recurrent neural network (RNN) that allows for better characterization of temporal trends, by learning various statistics of the input at different time scales. In our application, the complete states (injection rate, pressure and cumulative injection) at injectors and pressure states at producers are fed to SRU as the input and the phase rates at producers are treated as the output. Once the SRU is trained and validated, it is then used to assess the connectivity of each injector to any producer using permutation variable importance method, wherein inputs corresponding to an injector are shuffled and the increase in prediction error at a given producer is recorded as the importance (connectivity metric) of the injector to the producer. This method is tested in both synthetic and field-scale cases. The validation of the proposed data-driven inter-well connectivity assessment is performed using synthetic data from simulation models where inter-well connectivity can be easily measured using the streamline-based flux allocation. The SRU model is shown to offer excellent prediction performance on the synthetic case. Despite significant measurement noise and frequent well shut-ins imposed in the field-scale case, the SRU model offers good prediction accuracy, the overall relative error of the phase production rates at most producers ranges from 10% to 30%. It is shown that the dominant connections identified by the data-driven method and streamline method are in close agreement. This significantly improves confidence in our data-driven procedure. The novelty of this work is that it is purely data-driven method and can directly interpret routine surface measurements to intuitive subsurface knowledge. Furthermore, the streamline-based validation procedure provides physics-based backing to the results obtained from data analytics. This study results in a reliable and efficient data analytics framework that is well-suited for large field applications.

42 ENGINEERING↗

Denoising diffusion models with geometry adaptation for high fidelity calorimeter simulation

Simulation is crucial for all aspects of collider data analysis, but the available computing budget in the High Luminosity LHC era will be severely constrained. Generative machine learning models may act as surrogates to replace physics-based full simulation of particle detectors, and diffusion models have recently emerged as the state of the art for other generative tasks. Here, we introduce CaloDiffusion, a denoising diffusion model trained on the public CaloChallenge datasets to generate calorimeter showers. Our algorithm employs 3D cylindrical convolutions, which take advantage of symmetries of the underlying data representation. To handle irregular detector geometries, we augment the diffusion model with a new geometry latent mapping (GLaM) layer to learn forward and reverse transformations to a regular geometry that is suitable for cylindrical convolutions. The showers generated by our approach are nearly indistinguishable from the full simulation, as measured by several different metrics.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Neural network–based pore flow field prediction in porous media using super resolution

Direct pore-scale simulations of fluid flow through porous media are computationally expensive to perform for realistic systems. Previous works have demonstrated using the geometry of the microstructure of porous media to predict the velocity fields therein based on neural networks. However, such trained neural networks do not perform well for unseen porous media with a large degree of heterogeneity. In this study we propose that incorporating a coarse velocity field in the input of neural networks is an effective way to improve the prediction performance. The coarse velocity field can be simulated with a low computational cost and provides global information to regularize the ill-posedness of the learning problem, which is usually caused by the use of local geometries due to the computational resource constraints. We show that incorporating the coarse-mesh velocity field significantly improves the prediction accuracy of the fine-mesh velocity field by comparison to the prediction that relies on geometric information alone, especially for the porous medium with a large interior vuggy pore space. We also show the flexibility of training the network in using coarse velocity fields with various resolutions. The results suggest that even using coarse velocity field with a very low resolution, the predictions are still enhanced and close to the ground truths. The feasibility of the method is further demonstrated by testing the trained network on real rocks. This study highlights the merits of incorporating a coarse-mesh velocity field into the input for neural networks, which provides global, physics-based information for the model, thereby improving the model's generalization capability.

42 ENGINEERING↗

A Multi-Disciplinary Analysis Framework for the Design of Small Launch Vehicles

Between the years of 1995 and 2014 the number of small satellites (1-500 kg) went up from 20 to 180. [1] Out of the 180 launched in 2014 66% were Nano satellites (1-10 kg). [1] With this trend of smaller satellites, one would expect a rise in number of small launch vehicles (SLVs are defined by capability to carry 1-100 kg to orbit), but this has not happened: [3] only 8% of all small satellites are launched on SLVs and the others become secondary payloads on regular launch vehicles. [1] This results in small satellites being placed into either suboptimal orbits or waiting for launch schedules to align with bigger launches, resulting in long waiting times. Dedicated SLVs could improve the responsiveness of small satellite launches, but SLV design is complicated by the large architecture space needed to be explored in order to find efficient and affordable designs. The SLV architecture trade space contains many discrete options, e.g. solid vs. liquid fuels, air launch vs. ground launch, number of stages, etc. [2][3] Performing detailed analysis for all the options at once would be prohibitive. Thus, a sizing environment/framework that is capable of providing necessary information for conceptual-level trade studies and can rapidly explore the vast SLV architecture space is necessary. The framework, illustrated in Figure 1, consists of four disciplines central to the sizing and synthesis of launch vehicles: propulsion, aerodynamics, structures, and trajectory. For the aerodynamics and trajectory disciplines, Missile DATCOM and POST2 are used, respectively. The propulsion and structures disciplines are represented in the framework with tools developed at ASDL Georgia Tech. For propulsion, the Solid Motor Analysis Code (SMAC) is a physics-based conceptual design tool for solid rocket motors. SMAC is capable of geometric burn simulation, ballistic analysis, and prediction of thrust performance. [4] For structures, Launch Vehicle Structural Analysis (LVSA) tool is a physics-based tool that focuses on structural dynamic analysis with sizing capability. These tools are integrated into the framework illustrated in Figure 1 with the corresponding connections described in Table 1. The process flow is as follows: first, SMAC sizes insulation and calculates maximum operating pressures for each vehicle stage. The MEOP and insulation thickness values from SMAC are fed to LVSA which then utilizes this information to size the motor casing. This creates a feedback loop between LVSA and SMAC that converges on the radius available for fuel, casing, and insulation thickness. Once the stage sizing is converged on SMAC creates an engine deck that is passed to POST2. Next, the data from SMAC and LVSA goes into Missile DATCOM which generates an aerodynamics database for POST2. Finally, POST2 performs a targeting optimization while maximizing the payload mass to orbit. Within the framework, POST2 is automated in order to be robust to a wide variety of possible designs by performing a Monte Carlo simulation over the initialization vector of the POST2 optimization variables. To demonstrate the capability of this framework, a sample problem of exploring the design space of an SLV capable of placing satellites into a low Earth orbit (inclination=47 deg, 196.5 by 211.3 nm) is used. This sample problem is a ground-launched SLV consisting of four in-line SRM stages. The multidisciplinary design analysis (MDA) environment is utilized to explore a design space consisting of 22 continuous and 12 discrete variables, shown by Table 1 by blocks 1,2,3. Running a full factorial design of experiments (DOE) would have been prohibitively expensive even with this reduced design space, thus a space filling design with 3,502 and then additional expansion of 2,602 cases was used. The first DOE consisted of 3,502 cases, and all of the variables were varied. These input variables are listed in blocks 1, 2, and 3 in Table 1. Most of the variables are propulsion related with stage lengths determining the delta-V split of the SLV. The expansion consisted of 2,602 cases, and the continuous variables were set to be equal to the most promising designs from the sizing of first DOE. For each set of continuous variables, the discrete variables (grain type, star points, and propellant) were to varied. The results of the DOE can be seen in Figure 2; each of the points in this plot represents a closed launch vehicle that reaches the targeted orbit. For each of the cases, there is information on flown trajectory, structural, and propulsion properties of the SLV. For example, Figure 3 shows changes in altitude and velocity with time for a particular case. The right side of Figure 3 clearly shows the coasting (slow decrease) and burning phases (sharp increase) of the SLV mission. LV mass is positively correlated with the optimized payload mass to orbit because heavier LVs carry more fuel and thus have more stored chemical energy. Furthermore, for any given payload mass to orbit, the most efficient design would result in the smallest LV. The results as visualized in Figure 2 shows this tradeoff, and the Pareto frontier of the efficient designs can be seen along the dotted line. Figure 2 can be divided into regions with the lowest mass vehicles corresponding to the minimum bound on radius, and the highest mass vehicles corresponding to the maximum bound on radius. Within a mass region, the discrete variables, such as propellant type and propellant grain arrangement, have the most effect on payload mass. This paper presents an MDA framework that can perform an automated physics-based sizing of SLV designs and a corresponding methodology to utilize the MDA to explore the design space of SLVs. The proposed methodology was applied to a perform a design space exploration for a four stage SLV. The outputs show the expected pareto frontier forming and provide detailed information about the SLV performance and staging. Using the produced data, it will be possible to select a set of pareto optimal designs that can then be further explored in subsequent design cycles. This demonstration shows that automated design space exploration should be used in the early phase design of future SLV concepts.

Nikita S Birbasov↗

Insights into Slip-Rate Time Functions, Rupture Parameter Correlations, and Ground Motions from Validated Multicycle Earthquake Ruptures

Earthquake strong-motion predictions using kinematic source modeling require knowledge of the slip-rate functions (SRFs) along the rupture and their distinct characteristics in asperities, background (off-asperity) areas and near the surface. Here, in this study we analyzed SRFs from well-validated, self-consistent, and fully dynamic rupture models from earthquake cycles obeying a rate-and-state friction law, from our companion study (Galvez et al., 2021). The shapes of SRFs in asperities are well described by the regularized Yoffe function (RYF), which has only two parameters: rise time T r and smoothing time T s , which control the generation of long- and short-period ground motions, respectively. In background areas, we demonstrate that, in addition to the primary rupture, multiple secondary ruptures may also nucleate from rupture heterogeneities related to asperities, resulting in SRFs with multiple peaks. Because it is impossible to fit a multiple-peak SRF by the single-peak RYF, we describe SRFs in background areas in an effective way by fitting their amplitude spectra with the RYF spectra. Such spectrally effective RYFs capture salient aspects of seismic-wave generation and can be used in rupture generators for strong motion prediction. We found that small T s values correlate with small characteristic weakening distances, large peak slip rates (PSRs), and large rupture velocities. T r values are larger in background areas and smaller in asperities. Within the shallow aseismic zone, T s values approximately quadruple whereas T r values approximately double. Because of this dominant T s increase, PSR values decrease in the near-surface zone. These features indicate that the generation of strong motions by the near-surface portions of the rupture is negligible in the studied scenarios.

Geosciences↗

Recurrent neural network-based multiaxial plasticity model with regularization for physics-informed constraints

We report a recurrent neural network (RNN) based model is developed as a surrogate to predict nonlinear plastic response under multiaxial loading. The RNN-based model is trained and tested on stress versus strain curves generated using a numerical solution based on the classical radial return method. Besides simply learning the basic constitutive relationship, a novel approach is taken to enforce certain physical conditions. Specifically, regularization is employed to maintain non-negative plastic power density throughout the loading history thereby ensuring monotonically increasing plastic work and thermodynamic consistency. Enforcing physics in this manner permits coupling of the data-driven RNN approach with physics-based knowledge and laws. This has the effect of reducing the necessary amount of data and ensuring known physical laws are not violated. Since, once trained, the model need not perform the expensive task of solving nonlinear equations, its efficiency is orders of magnitude greater than its numerical counterpart. The RNN-based model has been trained on varied sets of data and the accuracy on test datasets validated. The developed model is general and robust and has widespread application such as in the simulation of metal forming, large scale plasticity, and part life prediction.

42 ENGINEERING↗

Frame invariant neural network closures for Kraichnan turbulence

Numerical simulations of geophysical and atmospheric flows have to rely on parameterizations of subgrid scale processes due to their limited spatial resolution. Despite substantial progress in developing parameterization (or closure) models for subgrid scale (SGS) processes using physical insights and mathematical approximations, they remain imperfect and can lead to inaccurate predictions. In recent years, machine learning has been successful in extracting complex patterns from high-resolution spatio-temporal data, leading to improved parameterization models, and ultimately better coarse grid prediction. However, the inability to satisfy known physics and poor generalization hinders the application of these models for real-world problems. In this work, we put forth a frame invariant closure approach to improve the accuracy and generalizability of deep learning-based subgrid scale closure models by embedding physical symmetries directly into the structure of the neural network. Specifically, we utilized specialized layers within the convolutional neural network in such a way that desired constraints are theoretically guaranteed without the need for any regularization terms. We demonstrate our framework for a two-dimensional decaying turbulence test case mostly characterized by the forward enstrophy cascade. We show that our frame invariant SGS model (i) accurately predicts the subgrid scale source term, (ii) respects the physical symmetries such as translation, Galilean, and rotation invariance, and (iii) is numerically stable when implemented in coarse-grid simulation with generalization to different initial conditions and Reynolds number. This work opens up a possibility of connecting physics-based theories and data-driven modeling paradigms, and thus represents a promising step towards the development of physically consistent data-driven turbulence closure models.

42 ENGINEERING↗

Machine learning unifies flexibility and efficiency of spinodal structure generation for stochastic biomaterial design

Abstract Porous biomaterials design for bone repair is still largely limited to regular structures (e.g. rod-based lattices), due to their easy parameterization and high controllability. The capability of designing stochastic structure can redefine the boundary of our explorable structure–property space for synthesizing next-generation biomaterials. We hereby propose a convolutional neural network (CNN) approach for efficient generation and design of spinodal structure—an intriguing structure with stochastic yet interconnected, smooth, and constant pore channel conducive to bio-transport. Our CNN-based approach simultaneously possesses the tremendous flexibility of physics-based model in generating various spinodal structures (e.g. periodic, anisotropic, gradient, and arbitrarily large ones) and comparable computational efficiency to mathematical approximation model. We thus successfully design spinodal bone structures with target anisotropic elasticity via high-throughput screening, and directly generate large spinodal orthopedic implants with desired gradient porosity. This work significantly advances stochastic biomaterials development by offering an optimal solution to spinodal structure generation and design.

59 BASIC BIOLOGICAL SCIENCES↗

G2Aero (Separable shape tensors for aerodynamic design) [SWR-22-44]

G2Aero is a flexible and practical tool for design and deformation of 2D airfoils and 3D blades using data-driven approaches. G2Aero utilizes the geometry of matrix manifolds—specifically the Grassmannian—to build a novel framework for representing physics-based separable deformations of shapes. G2Aero offers the flexibility to generate perturbations in a customizable way over any portion of the blade. The G2Aero framework utilizes data-driven methods based on a curated database of physically relevant airfoils. Specific tools include: - principal geodesic analysis over normal coordinate neighborhoods of matrix manifolds; - a variety of data-regularized deformations to nominal 2D airfoil shapes; - Riemannian interpolation connecting a sequence of airfoil cross-sections to build 3D blades from 2D data; - consistent perturbations over the span of interpolated 3D blades based on dominant modes from the data-driven analysis.

Doronina, Olga↗