Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “turbulence 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 451 records · Page 25
All sourcesNASA NTRSDOE OSTI
DOE OSTI2022

Three-dimensional simulations of reshocked inclined Richtmyer-Meshkov instability: Effects of initial perturbations

The effect of initial perturbations on the evolution of the inclined Richtmyer-Meshkov turbulent mixing layer before and after reshock initiated by a shock wave with Mach number 1.55 is investigated through three-dimensional (3D) simulations using the flash code. The 3D simulations aim to reproduce both predominantly single-mode and multimode interfaces between light and heavy gases (N 2 -CO 2 , Atwood number, A≈0.22; amplitude to wavelength ratio of 0.088) which were created in an inclined shock tube facility to analyze the effects of initial conditions on mixing development in the entire flow field. The two-dimensional center slices of 3D simulations are compared with the experimental results to validate the computational code. Mixing width, mixed mass, mixed-mass thickness, and circulation in addition to concentration fields are shown to be in good agreement with the experimental data. The three-dimensional density and vorticity fields are first presented to qualitatively describe the flow behavior before and after reshock. Several measured density/velocity-related quantities indicate that the growth of the mixing material is strongly dependent on initial conditions. Before reshock and at early times after reshock, flow is clearly maintaining the memory of initial perturbations. However, at late time after reshock, although the large wavelength feature still dominates the flow motion, and the morphology of the two different interfaces indicates several differences, by breakdown of large-scale coherent structures to much finer scales, the memory of small scales of the multimode initial perturbation is not as clear as pre-reshock. Regarding three-dimensionality of the flow, before reshock in the multimode case, the baroclinic vorticity production, circulation, turbulent kinetic energy, and turbulent mass flux suggest that the small-scale roll-up features along the large inclined wavelength quickly evolves in all three dimensions. The coherent vortex tubes break down to smaller wormlike vortex structures, and turbulent fluctuations in the out-of-plane dimension are comparable to the spanwise direction. After reshock, this three-dimensionality of mixing growth was observed in the flow for both initial conditions. The results of this work represent a significant extension of previous computational studies performed on this specific topic. A different code with a different numerical method is validated through comparison with the experimental data. The initial perturbations are directly measured from the experimental results. Moreover, the entire three-dimensional experimental shock tube domain is simulated, and more quantities are investigated to understand the mixing mechanism and instability evolution in all three dimensions.

42 ENGINEERING↗
DOE OSTI2022

A two-fluid single-column model of turbulent shallow convection. Part II: Single-column model formulation and numerics

The two-fluid single-column model of Thuburn et al. (Quart. J. R. Meteorol. Soc., 2019, 145, 1535–1550) is extended to include moisture and horizontal wind shear. Turbulent kinetic energy is introduced as a prognostic variable, dependence on a diagnosed boundary-layer height is removed, and subfilter fluxes are approximated using a two-fluid version of a Mellor–Yamada scheme. Three mechanisms for entrainment and detrainment processes are introduced, which represent entrainment of unstable air at the surface, forced detrainment of air at the top of the boundary/cloud layers, and turbulent mixing that relaxes the convective fluid to a reference profile. A semi-implicit Eulerian discretization replaces the semi-implicit semi-Lagrangian implementation of Thuburn et al. (Quart. J. R. Meteorol. Soc., 2019, 145, 1535–1550) to improve numerical stability and conservation. The equations for the implicit time step are solved using a quasi-Newton method, which is shown to perform well in numerical tests for conservation and convergence. The two-fluid single-column model presented in this article will be applied to simulations of shallow cumulus convection in Part III.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2021

Encoder–decoder neural network for solving the nonlinear Fokker–Planck–Landau collision operator in XGC

An encoder–decoder neural network has been used to examine the possibility for acceleration of a partial integro-differential equation, the Fokker–Planck–Landau collision operator. This is part of the governing equation in the massively parallel particle-in-cell code XGC, which is used to study turbulence in fusion energy devices. The neural network emphasizes physics-inspired learning, where it is taught to respect physical conservation constraints of the collision operator by including them in the training loss, along with the ℓ 2 loss. In particular, network architectures used for the computer vision task of semantic segmentation have been used for training. A penalization method is used to enforce the ‘soft’ constraints of the system and integrate error in the conservation properties into the loss function. During training, quantities representing the particle density, momentum and energy for all species of the system are calculated at each configuration vertex, mirroring the procedure in XGC. This simple training has produced a median relative loss, across configuration space, of the order of 10 –4 , which is low enough if the error is of random nature, but not if it is of drift nature in time steps. The run time for the current Picard iterative solver of the operator is O(n 2 ), where n is the number of plasma species. As the XGC1 code begins to attack problems including a larger number of species, the collision operator will become expensive computationally, making the neural network solver even more important, especially since its training only scales as O(n). Here, a wide enough range of collisionality has been considered in the training data to ensure the full domain of collision physics is captured. An advanced technique to decrease the losses further will be subject of a subsequent report. Eventual work will include expansion of the network to include multiple plasma species.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2023

Study of the GeV to TeV morphology of the γ Cygni SNR (G 78.2+2.1) with MAGIC and Fermi-LAT

Context: Diffusive shock acceleration (DSA) is the most promising mechanism that accelerates Galactic cosmic rays (CRs) in the shocks of supernova remnants (SNRs). It is based on particles scattering caused by turbulence ahead and behind the shock. The turbulence upstream is supposedly generated by the CRs, but this process is not well understood. The dominant mechanism may depend on the evolutionary state of the shock and can be studied via the CRs escaping upstream into the interstellar medium (ISM). Aims: Previous observations of the γ Cygni SNR showed a difference in morphology between GeV and TeV energies. Since this SNR has the right age and is at the evolutionary stage for a significant fraction of CRs to escape, our aim is to understand γ-ray emission in the vicinity of the γ Cygni SNR. Methods: We observed the region of the γ Cygni SNR with the MAGIC Imaging Atmospheric Cherenkov telescopes between 2015 May and 2017 September recording 87 h of good-quality data. Additionally, we analysed Fermi-LAT data to study the energy dependence of the morphology as well as the energy spectrum in the GeV to TeV range. The energy spectra and morphology were compared against theoretical predictions, which include a detailed derivation of the CR escape process and their γ-ray generation. Results: The MAGIC and Fermi-LAT data allowed us to identify three emission regions that can be associated with the SNR and that dominate at different energies. Our hadronic emission model accounts well for the morphology and energy spectrum of all source components. It constrains the time-dependence of the maximum energy of the CRs at the shock, the time-dependence of the level of turbulence, and the diffusion coefficient immediately outside the SNR shock. Further, while in agreement with the standard picture of DSA, the time-dependence of the maximum energy was found to be steeper than predicted, and the level of turbulence was found to change over the lifetime of the SNR.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2021

Response of thermal and fast-ion transport to beam ion population, rotation and T e / T i in the DIII-D steady state hybrid scenario

The thermal and fast-ion transport properties of DIII-D steady-state hybrid discharges with normalized beta β N ≳ 3 are studied at low injected torque and an increased electron to ion temperature ratio T e / T i . Linear stability analysis performed with the TGLF turbulent code indicates that a high-k mode is usually dominant at smaller radii, whereas a low-k mode is usually dominant at larger radii in these plasmas. A reduction in the net injected torque from 8.6 to 4.3 N-m leads to reduced E × B shear and hence, an enhanced turbulence that was observed on the Doppler backscattering diagnostic and was also computed with TGLF. As T e / T i in the core was increased from 0.57 to 0.66 by adding electron cyclotron current drive (ECCD) to these plasmas, higher levels of transport are observed with increased high-k modes indicated by TGLF. The fast-ion transport level varied over an order of magnitude in these discharges depending on whether Alfvén eigenmodes, fishbones, or no instabilities were observed. Hybrid plasmas with fishbones have decreased fast-ion transport, compared to plasmas with Alfvén eigenmodes, since they are resonant with a smaller portion of phase space and their resonance is farther from the wall. This reduction in fast-ion transport with ECCD mitigates the increase in turbulent transport, resulting in higher performance than expected during strong electron heating. Similarly, the lowest fast-ion transport was observed in the low torque plasma, which also led to better than expected performance at this torque value. Finally, the thermal and fast-ion transport changes observed as the torque/rotation and T e / T i are varied indicate possible methods for transferring this scenario to a reactor.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2024

CFD simulations of Molten Salt Fast Reactor core cavity flows

Computational Fluid Dynamics (CFD) has become increasingly important in the research and development of advanced nuclear reactors. Here, in the current study, extensive CFD simulations were conducted for the coolant flow in Molten Salt Fast Reactor (MSFR) core models using the state-of-the-art spectral element flow solver Nek5000 and multiscale coarse-mesh thermal-hydraulic software Pronghorn. The underlying motivation is to seek an in-depth understanding of how the internal velocity distribution can be influenced by the MSFR core cavity shape, the Reynolds number, turbulence modeling options and the inlet boundary conditions. The CFD techniques involved in this investigation range from coarse-mesh CFD, RANS modeling, to the high-fidelity LES calculations. Specifically, a series of RANS simulations were performed for the 2-D axisymmetric core model and 3-D wedge domains to study the flow distribution inside the MSFR core. It is observed that a proper representation of the MSFR inlet channel duct is important for the prediction of internal flow distribution. It is also showcased here how researchers can leverage the Nek5000 CFD results to calibrate more efficient coarse-mesh CFD tools, like Pronghorn, for the actual MSFR design needs. Moreover, this paper highlights a 3-D LES model for an entire MSFR core using the spectral element method and demonstrates the feasibility of this modeling approach. The readiness and potential limitations of the RANS approach are examined with respect to the high-fidelity LES simulations. The present investigation lays a solid foundation as we are leveraging the high-fidelity CFD capabilities to inform MSFR design efforts.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2026

Wake-Resolving Acoustic Tomography: Advances through Numerical Covariance Methods

Acoustic tomography offers path-integrated measurements of atmospheric velocity and temperature fluctuations with high spatial resolution. Classical implementations of time-dependent stochastic inversion rely on homogeneous, isotropic covariance models that are poorly suited to the anisotropic structure of wind turbine wakes. By directly estimating heterogeneous covariances from large-eddy simulations (LESs) into the time-dependent stochastic inversion operator, we relax implicit assumptions in the analytical models used historically. Retrievals using these LES-informed models improve agreement with true fields in variance, turbulent kinetic energy, and spectral content compared to analytical and precursor-based covariance models. The results indicate that LES-informed covariance models can enhance the accuracy of acoustic tomography retrievals in complex, anisotropic flows such as wind turbine wakes in some cases and highlight instances where analytical models still offer competitive performance, despite their simplifying assumptions.

17 WIND ENERGY↗
DOE OSTI2021

Mapping the magnetic field in the Taurus/B211 filamentary cloud with SOFIA HAWC + and comparing with simulation

ABSTRACT Optical and infrared polarization mapping and recent Planck observations of the filametary cloud L1495 in Taurus show that the large-scale magnetic field is approximately perpendicular to the long axis of the cloud. We use the HAWC + polarimeter on SOFIA to probe the complex magnetic field in the B211 part of the cloud. Our results reveal a dispersion of polarization angles of 36°, about five times that measured on a larger scale by Planck. Applying the Davis–Chandrasekhar–Fermi (DCF) method with velocity information obtained from Institut de Radioastronomie Millimétrique 30 m C18O(1-0) observations, we find two distinct sub-regions with magnetic field strengths differing by more than a factor 3. The quieter sub-region is magnetically critical and sub-Alfv$\acute{\rm e}$nic; the field is comparable to the average field measured in molecular clumps based on Zeeman observations. The more chaotic, super-Alfv$\acute{\rm e}$nic sub-region shows at least three velocity components, indicating interaction among multiple substructures. Its field is much less than the average Zeeman field in molecular clumps, suggesting that the DCF value of the field there may be an underestimate. Numerical simulation of filamentary cloud formation shows that filamentary substructures can strongly perturb the magnetic field. DCF and true field values in the simulation are compared. Pre-stellar cores are observed in B211 and are seen in our simulation. The appendices give a derivation of the standard DCF method that allows for a dispersion in polarization angles that is not small, present an alternate derivation of the structure function version of the DCF method, and treat fragmentation of filaments.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2022

Super‐Droplet Method to Simulate Lagrangian Microphysics of Nuclear Fallout in a Homogeneous Cloud

Abstract Nuclear detonations produce hazardous local and global particles or fallout. Predicting fallout size, chemical components, and location is necessary to inform officials and determine immediate guidance for the public. However, existing nuclear detonation fallout models prescribe the particle size distributions based on limited observations. In this work, we apply the super‐droplet method, which is a numerical modeling technique developed for cloud microphysics, to simulate size distributions of particles in a mushroom cloud formed post‐detonation of a nuclear device. We model fallout formation and evolution with homogeneous nucleation and condensation of a single species and a Monte Carlo coagulation algorithm. We verify the numerical methods representing coagulation and condensation processes against analytical test problems. Additionally, we explore several scenarios for the integral system mass and yield in equivalent kilotons (kt) of TNT (trinitrotoluene). The fallout size distribution median diameter d pg follows a scaling law based on the integral system mass m v 0 kg and yield Y kt: nm. We test the effect of cloud turbulence, enhanced nucleation and growth, and vapor volatility with a sensitivity study. The range in median diameter predictions for simulations of historical tests performed over the Pacific encompass the measurements of particles sampled from the cloud caps. Predicted median particle size ranges up to 217, 123, 86, and 35 nm for historical tests with yields of 0.2, 0.7, 2, and 10 Mt, respectively. This work can be expanded in many different directions to build a more predictive model for fallout formation.

45 MILITARY TECHNOLOGY, WEAPONRY, AND NATIONAL DEF↗
DOE OSTI2020

Spontaneous Raman–LIF–CO–OH measurements of species concentration in turbulent spray flames

This paper presents new measurements of species concentrations, temperature and mixture fraction in selected regions of a turbulent ethanol spray flame. The line-Raman–LIF–COsingle bondOH setup developed at the Sandia's Combustion Research Facility is utilised to probe regions of a spray flame where laser breakdown of liquid droplets is avoided and the remaining interferences can be corrected. The spray flame is stabilised on the piloted Sydney needle spray burner, where axial translation of the liquid injecting needle in the air-blast stream can transition the spray from dilute to dense. The solution to obtaining successful measurements is found to be multifaceted and includes: the appropriate selection of flame conditions; high sensitivity of the Raman detection system permitting reduced laser energies; development of a pre-processing algorithm to reject strong droplet interferences; and application of the hybrid matrix inversion method combined with wavelet denoising to account for interference corrections and noise at the very low signal levels obtained. Unique and necessary for the successful measurements reported in this paper, a pre-processing algorithm is outlined that removes data points corrupted with strong interferences from droplets. These interferences arise from a range of sources, but the most intense are due to the laser interaction with surrounding mist or liquid fragments, such that measurements near the jet centreline are corrupted and hence discarded. Reliable measurements of mixture fraction, temperature obtained from the sum of the species number densities, and species mole fractions are reported for regions in the flames sufficiently far from the centreline. The paper demonstrates the feasibility of the judicious use of Raman scattering in turbulent spray flames, the results of which will be extremely useful for validating numerical simulations.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2020

Impact of boundary layer simulation on predicting radioactive pollutant dispersion: a case study for HANARO Research Reactor using the WRF-MMIF-CALPUFF modeling system

Wind plays an important role in cases of unexpected radioactive pollutant dispersion, deciding distribution and concentration of the leaked substance. The accurate prediction of wind has been challenging in numerical weather prediction models, especially near the surface because of the complex interaction between turbulent flow and topographic effect. As such, in this study, we investigated the characteristics of atmospheric dispersion of radioactive material (i.e. 137 Cs) according to the simulated boundary layer around the HANARO research nuclear reactor in Korea using the Weather Research and Forecasting (WRF)-Mesoscale Model Interface (MMIF)-California Puff (CALPUFF) model system. We examined the impacts of orographic drag on wind field, stability calculation methods, and planetary boundary layer parameterizations on the dispersion of radioactive material under a radioactive leaking scenario. We found that inclusion of the orographic drag effect in the WRF model improved the wind prediction most significantly over the complex terrain area, leading the model system to estimate the radioactive concentration near the reactor more conservatively. We also emphasized the importance of the stability calculation method and employing the skillful boundary layer parameterization to ensure more accurate low atmospheric conditions, in order to simulate more feasible spatial distribution of the radioactive dispersion in leaking scenarios.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2026

Benchmarking of three DWM-based wake models at below-rated wind speeds

Wind turbine wake models are essential tools for predicting power losses and structural loads in wind farms. Among these, the dynamic wake meandering (DWM) model, included as a recommended approach in the International Electrotechnical Commission design standard, is a widely used engineering-fidelity method that balances accuracy and computational cost. This study compares the performance of three DWM-based wake model implementations (from the Technical University of Denmark, the National Renewable Energy Laboratory, and the Institute for Energy Technology) under below-rated wind speed conditions. Model predictions of wake flow, power output, and structural loads for a four-turbine row are evaluated across different ambient turbulence levels and wind-direction misalignments and compared against high-fidelity large-eddy simulation results. All three models captured the overall wake evolution and mean turbine performance with reasonable accuracy; their predicted time-averaged thrust and power were typically within 5 %–10 % of the large-eddy simulation benchmark. However, notable differences emerged in wake structure and unsteady load predictions, with discrepancies increasing for turbines further downstream. These differences highlight the importance of modelling choices such as wake summation and turbulence treatment, which strongly influence power-deficit and fatigue-load predictions. Comparison with large-eddy simulations reveals each approach's strengths and weaknesses, indicating where improvements are needed. Overall, the findings point to specific refinements for DWM models to improve their fidelity, ultimately enabling more robust wake predictions for wind farm design and operation.

17 WIND ENERGY↗
DOE OSTI2024

Validation of the stochastic inversion algorithm for acoustic travel-time tomography: a large eddy simulation study

Acoustic tomography (AT) is explored as a remote sensing technique to obtain instantaneous snapshots of temperature and velocity fluctuations for wind energy applications. This study integrates Large Eddy Simulation (LES) with the Stochastic Inversion (SI) method to validate the algorithm’s capacity for accurate reconstruction of atmospheric fluctuations. The initial findings demonstrate the efficacy of the method in accurately capturing the predominant flow structures. Normalized L2 error evaluations further inform the algorithm’s precision, with errors accentuated in less sampled peripheral regions. The results underscore the method’s promise as a non-intrusive observational tool, with ongoing development poised to improve its precision and reliability.

17 WIND ENERGY↗
DOE OSTI2026

Estimating and Evaluating Roughness Length and Displacement Height in Heterogeneous Urban Environments

The roughness length (z 0 ) and displacement height (z d ) are essential surface-layer parameters in numerical models (e.g., weather, climate, wall-modeled LES, etc.). This work evaluates the consistency of z 0 and z d estimates from morphometric and anemometric methods using data from two eddy-covariance flux towers (AmeriFlux US-INg and US-INc) in Indianapolis, IN. Results show inconsistencies in estimated z 0 and z d values depending on the chosen method. The two evaluated anemometric methods estimate non-physical values of z d when compared to roughness elements surrounding both towers. Additionally, predictions of mean wind speed using surface-layer similarity theory with morphometric estimates exhibit a bias during near-neutral and stable conditions relative to observations. The overestimation of mean wind speed by surface layer similarity theory is consistent with previous observational and modeling studies in urban areas, suggesting that the application of similarity theories to urban environments may have limitations. Differentiation of vegetation from built structures appears to impact morphometric z 0 and z d estimates, particularly where vegetation is abundant; however, it has little impact on correcting biases in the similarity theory. Specifically, we find that existing similarity theories using morphometric estimates underestimate integral velocity and length scales, and the degree of underestimation depends on the stability conditions. Accounting for the degree of anisotropy in surface-layer turbulence helps reduce the biases between similarity theories and observations during unstable conditions, but not in near-neutral cases. Future work is needed to identify the cause of such biases for near-neutral conditions.

Aerodynamic roughness length↗
DOE OSTI2024

Optimization of nonlinear turbulence in stellarators

We present new stellarator equilibria that have been optimized for reduced turbulent transport using nonlinear gyrokinetic simulations within the optimization loop. The optimization routine involves coupling the pseudo-spectral GPU-native gyrokinetic code GX with the stellarator equilibrium and optimization code DESC . Since using GX allows for fast nonlinear simulations, we directly optimize for reduced nonlinear heat fluxes. To handle the noisy heat flux traces returned by these simulations, we employ the simultaneous perturbation stochastic approximation (SPSA) method that only uses two objective function evaluations for a simple estimate of the gradient. We show several examples that optimize for both reduced heat fluxes and good quasi-symmetry as a proxy for low neoclassical transport. Finally, we run full transport simulations using the T3D stellarator transport code to evaluate the changes in the macroscopic profiles.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2023

Tidal Turbine Benchmarking Project: Stage I - Steady Flow Blind Predictions: Preprint

This paper presents the first blind prediction stage of the Tidal Turbine Benchmarking Project being conducted and funded by the UK's EPSRC and Supergen ORE Hub. In this first stage, only steady flow conditions, at low and elevated turbulence (3.1%) levels, were considered. Prior to the blind prediction stage, a large laboratory scale experiment was conducted in which a highly instrumented 1.6m diameter tidal rotor was towed through a large towing tank in well-defined flow conditions with and without an upstream turbulence grid. Details of the test campaign and rotor design were released as part of this community blind prediction exercise. Participants were invited to use a range of engineering modelling approaches to simulate the performance and loads of the turbine. 26 submissions were received from 12 groups from across academia and industry using solution techniques ranging from blade resolved computational fluid dynamics through actuator line, boundary integral element methods, vortex methods to engineering Blade Element Momentum methods. The comparisons between experiments and blind predictions were extremely positive helping to provide validation and uncertainty estimates for the models, but also validating the experimental tests themselves. The exercise demonstrated that the experimental turbine data provides a robust data set against which researchers and design engineers can test their models and implementations to ensure robustness in their processes, helping to reduce uncertainty and provide increased confidence in engineering processes. Furthermore, the data set provides the basis by which modellers can evaluate and refine approaches.

benchmarking↗
DOE OSTI2021

A vortex sheet based analytical model of the curled wake behind yawed wind turbines

Motivated by the need for compact descriptions of the evolution of non-classical wakes behind yawed wind turbines, we develop an analytical model to predict the shape of curled wakes. Interest in such modelling arises due to the potential of wake steering as a strategy for mitigating power reduction and unsteady loading of downstream turbines in wind farms. We first estimate the distribution of the shed vorticity at the wake edge due to both yaw offset and rotating blades. By considering the wake edge as an ideally thin vortex sheet, we describe its evolution in time moving with the flow. Vortex sheet equations are solved using a power series expansion method, and an approximate solution for the wake shape is obtained. The vortex sheet time evolution is then mapped into a spatial evolution by using a convection velocity. Apart from the wake shape, the lateral deflection of the wake including ground effects is modelled. Our results show that there exists a universal solution for the shape of curled wakes if suitable dimensionless variables are employed. For the case of turbulent boundary layer inflow, the decay of vortex sheet circulation due to turbulent diffusion is included. Finally, we modify the Gaussian wake model by incorporating the predicted shape and deflection of the curled wake, so that we can calculate the wake profiles behind yawed turbines. Model predictions are validated against large-eddy simulations and laboratory experiments for turbines with various operating conditions.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2023

Magnetohydrodynamics with physics informed neural operators

Abstract The modeling of multi-scale and multi-physics complex systems typically involves the use of scientific software that can optimally leverage extreme scale computing. Despite major developments in recent years, these simulations continue to be computationally intensive and time consuming. Here we explore the use of AI to accelerate the modeling of complex systems at a fraction of the computational cost of classical methods, and present the first application of physics informed neural operators (NOs) (PINOs) to model 2D incompressible magnetohydrodynamics (MHD) simulations. Our AI models incorporate tensor Fourier NOs as their backbone, which we implemented with the TensorLY package. Our results indicate that PINOs can accurately capture the physics of MHD simulations that describe laminar flows with Reynolds numbers Re ⩽ 250 . We also explore the applicability of our AI surrogates for turbulent flows, and discuss a variety of methodologies that may be incorporated in future work to create AI models that provide a computationally efficient and high fidelity description of MHD simulations for a broad range of Reynolds numbers. The scientific software developed in this project is released with this manuscript.

97 MATHEMATICS AND COMPUTING↗