SEARCH · Engineering Papers

Results for “Diffusion equation”

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.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 739 records · Page 41

All sources NASA NTRS DOE OSTI

DOE OSTI2023

A finite element method for angular discretization of the radiation transport equation on spherical geodesic grids

Discrete ordinate (S N ) and filtered spherical harmonics (FP N ) based schemes have been proven to be robust and accurate in solving the Boltzmann transport equation but they have their own strengths and weaknesses in different physical scenarios. We present a new method based on a finite element approach in angle that combines the strengths of both methods and mitigates their disadvantages. The angular variables are specified on a spherical geodesic grid with functions on the sphere being represented using a finite element basis. A positivity-preserving limiting strategy is employed to prevent non-physical values from appearing in the solutions. Here, the resulting method is then compared with both S N and FP N schemes using four test problems and is found to perform well when one of the other methods fail.

79 ASTRONOMY AND ASTROPHYSICS↗

DOE OSTI2021

A Hybrid Finite-Volume, Discontinuous Galerkin Discretization for the Radiative Transport Equation

In this report we propose a hybrid spatial discretization for the radiative transport equation that combines a second-order discontinuous Galerkin (DG) method and a second-order finite-volume (FV) method. The strategy relies on a simple operator splitting that has been used previously to combine different angular discretizations. Unlike standard FV methods with upwind fluxes, the hybrid approach is able to accurately simulate problems in scattering dominated regimes. However, it requires less memory and yields a faster computational time than a uniform DG discretization. In addition, the underlying splitting allows naturally for hybridization in both space and angle. Numerical results are given to demonstrate the efficiency of the hybrid approach in the context of discrete ordinate angular discretizations and Cartesian spatial grids.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2023

MPACT 4.4 Theory Manual

MPACT is a three-dimensional (3D) full-core neutron transport code capable of calculating subpin power distributions. Calculations are based on the Boltzmann transport equation for neutron fluxes for problems in which the detailed geometrical configuration of fuel components such as the pellet and cladding are explicitly retained. The cross-section data needed for the neutron transport calculation are obtained directly from a multigroup cross section library, which has traditionally been used by lattice physics codes to generate few-group homogenized cross sections for nodal core simulators. Hence, MPACT assumes neither a priori homogenization nor group condensation for the full core spatial solution. The 3D MPACT transport solution can be obtained using the method of characteristics (MOC), which employs discrete ray tracing within each fuel pin. However, for practical reactor applications, the direct application of MOC to 3D core configurations requires an excessive amount of memory and computing time due to the very large number of rays. For practical 3D full-core calculations, MPACT commonly uses an approximate “2D/1D” method that treats the radial (x and y) variables differently from the axial (z) variable. In particular, the radial dependence of the solution is calculated using transport theory, and the axial dependence is calculated using diffusion or P 3 theory. The 2D/1D method requires the core to be divided into a vertical stack of axial slices with a thickness of Δ z ≈ 5–10 cm. Each axial slice is divided radially into coarse spatial cells with boundaries that usually constitute the pin cell boundaries, for which Δ x = Δ y ≈ 1.5 cm. Then, each coarse radial cell (pin cell) is divided into 50–100 fine radial cells, which resolve the angular flux in the fuel, cladding, and moderator regions.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

DOE OSTI2023

Hierarchical Model Reduction Driven by Machine Learning for Parametric Advection-Diffusion-Reaction Problems in the Presence of Noisy Data

Abstract We propose a new approach to generate a reliable reduced model for a parametric elliptic problem, in the presence of noisy data. The reference model reduction procedure is the directional HiPOD method, which combines Hierarchical Model reduction with a standard Proper Orthogonal Decomposition, according to an offline/online paradigm. In this paper we show that directional HiPOD looses in terms of accuracy when problem data are affected by noise. This is due to the interpolation driving the online phase, since it replicates, by definition, the noise trend. To overcome this limit, we replace interpolation with Machine Learning fitting models which better discriminate relevant physical features in the data from irrelevant unstructured noise. The numerical assessment, although preliminary, confirms the potentialities of the new approach.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2024

BCS-BEC crossover of the strongly interacting 6 Li − 40 K mixture

Here, we present quantum Monte Carlo calculations of the properties of a two-component mass-imbalanced Fermi gas, corresponding to the 6 Li − 40 K mixture. We compute the equation of state of the unpolarized system as a function of the scattering length with particular attention paid to the unitary limit, where the effect of the effective range of the interaction is explored. In order to investigate differences from the equal-mass case we also compute the pair-distribution function and the momentum distribution over a range of interaction strengths, which can provide information about the structure of the system. Finally, we compute the heavy-light quasiparticle spectrum for the full crossover regime. Our theoretical predictions, based on quantum Monte Carlo calculations, should inform future theoretical studies and can be tested by experiments with ultracold fermionic gases.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

DOE OSTI2023

Energetic Electron Precipitation Driven by Electromagnetic Ion Cyclotron Waves from ELFIN’s Low Altitude Perspective

We review comprehensive observations of electromagnetic ion cyclotron (EMIC) wave-driven energetic electron precipitation using data collected by the energetic electron detector on the Electron Losses and Fields InvestigatioN (ELFIN) mission, two polar-orbiting low-altitude spinning CubeSats, measuring 50-5000 keV electrons with good pitch-angle and energy resolution. EMIC wave-driven precipitation exhibits a distinct signature in energy-spectrograms of the precipitating-to-trapped flux ratio: peaks at >0.5 MeV which are abrupt (bursty) (lasting ~17 s, or ΔL ~ 0.56) with significant substructure (occasionally down to sub-second timescale). We attribute the bursty nature of the precipitation to the spatial extent and structuredness of the wave field at the equator. Multiple ELFIN passes over the same MLT sector allow us to study the spatial and temporal evolution of the EMIC wave - electron interaction region. Case studies employing conjugate ground-based or equatorial observations of the EMIC waves reveal that the energy of moderate and strong precipitation at ELFIN approximately agrees with theoretical expectations for cyclotron resonant interactions in a cold plasma. Using multiple years of ELFIN data uniformly distributed in local time, we assemble a statistical database of ~50 events of strong EMIC wave-driven precipitation. Most reside at L ~ 5 – 7 at dusk, while a smaller subset exists at L ~ 8 – 12 at post-midnight. The energies of the peak-precipitation ratio and of the half-peak precipitation ratio (our proxy for the minimum resonance energy) exhibit an L-shell dependence in good agreement with theoretical estimates based on prior statistical observations of EMIC wave power spectra. The precipitation ratio’s spectral shape for the most intense events has an exponential falloff away from the peak (i.e., on either side of ~ 1.45 MeV). It too agrees well with quasi-linear diffusion theory based on prior statistics of wave spectra. It should be noted though that this diffusive treatment likely includes effects from nonlinear resonant interactions (especially at high energies) and nonresonant effects from sharp wave packet edges (at low energies). Sub-MeV electron precipitation observed concurrently with strong EMIC wave-driven >1 MeV precipitation has a spectral shape that is consistent with efficient pitch-angle scattering down to ~ 200-300 keV by much less intense higher frequency EMIC waves at dusk (where such waves are most frequent). At ~100 keV, whistler-mode chorus may be implicated in concurrent precipitation. These results confirm the critical role of EMIC waves in driving relativistic electron losses. Nonlinear effects may abound and require further investigation.

79 ASTRONOMY AND ASTROPHYSICS↗

DOE OSTI2020

Exploring an ‘ideal hill': how lithology and transport mechanisms affect the possibility of a steady state during weathering and erosion

We introduce a model of chemical reaction within hills to explore how evolving porosity (and by inference, permeability) affects flow pathways and weathering. The model consists of hydrologic and reactive-transport equations that describe alteration of ferrous minerals and feldspar. These reactions were chosen because previous work emphasized that oxygen- and acid-driven weathering affects porosity differently in felsic and mafic rocks. A parameter controlling the order of the fronts is presented. In the absence of erosion, the two reaction fronts move at different velocities and the relative depths depend on geochemical conditions and starting composition. In turn, the fronts and associated changes in porosity drastically affect both the vertical and lateral velocities of water flow. For these instances, estimates of weathering advance rates based on simple models that posit unidirectional constant-velocity advection do not apply. In the model hills, weathering advance rates diminish with time as the Darcy velocities decrease with depth. The system can thus attain a dynamical steady state at any erosion rate where the regolith thickness is constant in time and velocities of both fronts become equal to one another and to the erosion rate. The slower the advection velocities in a system, the faster it attains a steady state. For example, a massive rock with relatively fast-dissolving minerals such as diabase -- where solute transport across the reaction front mainly occurs by diffusion -- can reach a steady state more quickly than granitoid rocks in which advection contributes to solute transport. The attainment of a steady state is controlled by coupling between weathering and hydrologic processes that force water to flow horizontally above reaction fronts where permeability changes rapidly with depth and acts as a partial barrier to fluid flow.

54 ENVIRONMENTAL SCIENCES↗

DOE OSTI2024

A Fast Algebraic Multigrid Solver and Accurate Discretization for Highly Anisotropic Heat Flux I: Open Field Lines

We present a novel solver technique for the anisotropic heat flux equation, aimed at the high level of anisotropy seen in magnetic confinement fusion plasmas. Such problems pose two major challenges: (i) discretization accuracy and (ii) efficient implicit linear solvers. We simultaneously address each of these challenges by constructing a new finite element discretization with excellent accuracy properties, tailored to a novel solver approach based on algebraic multigrid (AMG) methods designed for advective operators. We pose the problem in a mixed formulation, introducing the directional temperature gradient as an auxiliary variable. The temperature and auxiliary fields are discretized in a scalar discontinuous Galerkin space with upwinding principles used for discretizations of advection. We demonstrate the proposed discretization’s superior accuracy over other discretizations of anisotropic heat flux, achieving error 1000x smaller for anisotropy ratio of 10 9 , for closed field lines. The block matrix system is reordered and solved in an approach where the two advection operators are inverted using AMG solvers based on approximate ideal restriction, which is particularly efficient for upwind discontinuous Galerkin discretizations of advection. To ensure that the advection operators are nonsingular, in this paper we restrict ourselves to considering open (acyclic) magnetic field lines for the linear solvers. We demonstrate fast convergence of the proposed iterative solver in highly anisotropic regimes where other diffusion-based AMG methods fail.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2022

Stabilizing and Improving Qubit Coherence by Engineering the Noise Spectrum of Two-Level Systems

Superconducting circuits are a leading platform for quantum computing. However, their coherence times are still limited and exhibit temporal fluctuations. Those phenomena are often attributed to the coupling between qubits and material defects that can be well described as an ensemble of two-level systems (TLSs). Among them, charge fluctuators inside amorphous oxide layers contribute to both low-frequency 1 / f charge noise and high-frequency dielectric loss, causing fast qubit dephasing and relaxation. Moreover, spectral diffusion from mutual TLS interactions varies the noise amplitude over time, fluctuating the qubit lifetime. Here, we propose to mitigate those harmful effects by engineering the relevant TLS noise spectral densities. Specifically, our protocols smooth the high-frequency noise spectrum and suppress the low-frequency noise amplitude via depolarizing and dephasing the TLSs, respectively. As a result, we predict a drastic stabilization in qubit lifetime and an increase in qubit pure dephasing time. Our detailed analysis of feasible experimental implementations shows that the improvement is not compromised by spurious coupling from the applied noise to the qubit.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

DOE OSTI2024

A Free-Standing Boron-Doped Diamond Grid Electrode for Fundamental Spectroelectrochemistry

Spectroelectrochemistry (SEC) is a powerful technique that enables a variety of redox properties to be studied, including formal potential (E o ), thermodynamic values (ΔG, ΔH, ΔS), diffusion coefficient (D), electron transfer stoichiometry (n), and others. SEC requires an electrode which light can pass through while maintaining sufficient electrical conductivity. This has been traditionally composed of metal or metal oxide films atop transparent substrates like glass, quartz, or metallic mesh. Robust electrode materials like boron-doped diamond (BDD) could help expand the environments in which SEC can be performed, but most designs are limited to thin films (~100–200 nm) on transparent substrates less resilient than free-standing BDD. Here, this work presents a free-standing BDD grid electrode (G-BDD) for fundamental SEC measurements, using the well-characterized Fe(CN) 6 3–/4– redox couple as proof-of-concept. With a combination of cyclic voltammetry (CV), thin-layer SEC, and chronoabsorptometry, several of the redox properties mentioned above were calculated and compared. For E o' , n, and D, similar results were obtained when comparing the CV [E o' = +0.279 (±0.002) V vs Ag/AgCl; n = 0.97; D = 4.1 × 10 –6 cm 2 ·s –1 ] and SEC [E o' = +0.278 (±0.001) V vs Ag/AgCl; n = 0.91; D = 5.2 × 10 –6 cm 2 ·s –1 ] techniques. Both values align with what has been previously reported. To calculate D from the SEC data, modification of the classical equation used in chronoabsorptometry was required to accommodate the G-BDD electrode geometry. Overall, this work expands on the applicability of SEC techniques and BDD as a versatile electrode material.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

DOE OSTI2026

Investigating shock-induced chemical reactions in Ni+Al multilayers: A continuum-based mesoscale approach with Arrhenius kinetics and artificial thermal conduction

A new continuum-based mesoscale modeling approach for shock-induced chemical reactions (SICRs) in Ni+Al multilayers is demonstrated in Sandia’s shock physics hydrocode, CTH. The approach utilizes Arrhenius-type kinetics and artificial thermal conduction. Our work builds upon previous efforts to parameterize equations of state for Ni x Al y [J. Appl. Phys. 137, 075102 (2025)], as well as simulations of inert shocks in realistic 2D microstructures [J. Appl. Phys. 137, 225301 (2025)]. To calibrate the reaction kinetics, pairs of the reaction coordinate, R′, vs time are extracted from the molecular dynamics (MD) literature. Here, the MD-informed kinetics are used to simulate the dynamic evolution of pressure and temperature in 2D mesoscale simulations. Overall, the MD-informed kinetics obtained for planar interfaces are too slow, as initial reaction is not observed on a nanosecond time scale. Even with quasi-periodic shock focusing leading to the formation of so-called “hot-spots,” the hot spots are unable to grow and coalesce using the fitted Arrhenius rate constants. However, by increasing the rate constants by two orders of magnitude, SICRs are observed at a shock pressure near 30 GPa, which is supported by experiments. Consequently, these mesoscale simulations suggest that unresolved shear-based mechanical mixing might possibly account for the discrepancies in kinetic rates, with shock-generated intense perturbations, interfacial vortical flows, and elevated temperatures serving as favorable reaction conditions. Future work will calibrate a shear-dependent reaction rate from the MD simulations with realistic interfaces that are wavy, diffuse, and disordered.

Kittell, David E. [Sandia National Laboratories (S↗

DOE OSTI2020

Improvements to a class of hybrid methods for radiation transport: Nyström reconstruction and defect correction methods

In this study, two modifications are introduced for improving the accuracy, versatility, and robustness of a class of hybrid methods for radiation transport. In general, such methods are constructed by splitting the radiative flux into collided and uncollided components to which low- and high-resolution angular approximations are applied, respectively. In this work we focus on discrete ordinates discretizations of high and low order. The first modification we introduce changes the way in which the collided component is mapped into the uncollided component at the end of each time step in a simulation. The new mapping is a Nyström-type reconstruction that is applicable to arbitrary discrete ordinates quadratures, is guaranteed to preserve positivity of the solution provided that all ordinate weights are positive, is significantly more accurate than previous methods, and can be readily extended to other discretizations such as moment methods, finite element methods, and diffusion approximations. The second modification leverages integral deferred correction (IDC) to iteratively correct for the splitting error introduced by the inconsistency in angular discretization between the collided and uncollided components, in addition to improving the accuracy of the low-order temporal error that is treated by traditional IDC methods. Numerical tests in one- and two-dimensional geometries are used to demonstrate the increased accuracy and efficiency of the proposed modifications. It is found that the two techniques combined yield methods with solution accuracy and memory requirements comparable to that of monolithic discrete ordinates methods while reducing runtime by as much as a factor of between two and ten, depending on the problem.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

DOE OSTI2023

MATBOX, an Open-Source Microstructure Analysis Toolbox for Meshing, Generation, Segmentation, and Characterization of 3D Heterogenous Volumes

Battery performance is strongly correlated with electrode microstructural properties. To account for its impact, lithium-ion battery (LIB) models either abstract the microstructural heterogeneity of composite electrodes using effective macroscopic properties (macro- or meso- scale models) or directly solve the system of equations on the microstructure geometry or mesh (microstructure-scale models). Therefore, to be adequate, both families of models require information from the microstructure geometry, which can be provided by the numerical tool presented in this work. MATBOX is a MATLAB open-source application [1] developed by NREL for performing various microstructure-related tasks including microstructure numerical generation, image filtering and microstructure segmentation, microstructure characterization and correlation, visualization, and microstructure meshing. MATBOX was originally developed for the analysis of LIB electrode microstructures; however, the algorithms provided by the toolbox are widely applicable to other heterogeneous materials. The toolbox provides a user-friendly experience thanks to a Graphical-User Interface, requires no coding by the user, and is well documented. This presentation will illustrate various MATBOX features for the characterization of a LIB electrode, including a fully automated Representative Volume Element (RVE) analysis, the numerical generation of complex 'virtual' microstructure, including dual-layer electrodes and carbon-binder additive phase, and the meshing of a complex NMC/graphite full cell microstructure suitable for 3D finite-element modeling. Other modules (segmentation, visualization, and correlation) will be briefly presented. Thanks to its modular, open-source approach, MATBOX can easily incorporate third-party algorithms to eventually build a standard in the field that will benefit the whole scientific community. Effective diffusion coefficient [2], additive phase numerical generation [3], and meshing [4] third-party algorithms have been already integrated in the toolbox with more to come.

DIRECT ENERGY CONVERSION,MATHEMATICS AND COMPUTING↗

DOE OSTI2021

A Kinetic Monte Carlo Approach for Simulating Cascading Transmission Line Failure

In this work, cascading transmission line failures are studied through a dynamical model of the power system operating under fixed conditions. The power grid is modeled as a stochastic dynamical system where first-principles electromechanical dynamics are excited by small Gaussian disturbances in demand and generation around a specified operating point. In this context, a single line failure is interpreted in a large deviation context as a first escape event across a surface in phase space defined by line security constraints. The resulting system of stochastic differential equations admits a transverse decomposition of the drift, which leads to considerable simplification in evaluating the quasipotential (rate function) and, consequently, computation of exit rates. Tractable expressions for the rate of transmission line failure in a restricted network are derived from large deviation theory arguments and validated against numerical simulations. Extensions to realistic settings are considered, and individual line failure models are aggregated into a Markov model of cascading failure inspired by chemical kinetics. Cascades are generated by traversing a graph composed of weighted edges representing transitions to degraded network topologies. Numerical results indicate that the Markov model can produce cascades with qualitative power-law properties similar to those observed in empirical cascades.

Kinetic Monte Carlo, rare events, Cascading Failur↗

DOE OSTI2020

Quantum Monte Carlo benchmarking of large noncovalent complexes in the L7 benchmark set

We have used diffusion Monte Carlo (DMC) to perform calculations on the L7 benchmark set. DMC is a stochastic numerical integration scheme in real-space and part of a larger set of quantum Monte Carlo methods. The L7 set was designed to test the ability of electronic structure methods to include dispersive interactions. While the agreement between DMC and quantum-chemical state-of-the-art methods is excellent for some of the structures, there are significant differences in others. In contrast to wavefunction-based quantum chemical methods, DMC is a first-principle many-body method with the many-body wavefunction evolving in real space. It includes explicitly all electron–electron interactions and is relatively insensitive to the size of the basis set.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

DOE OSTI2024

MOSCATO Development and Integration in Fiscal Year 2024

MOSCATO (Molten Salt Chemistry and Transport) is a multiphysics code that provides high-fidelity, coupled simulations of fluid flow, heat transfer, mass transfer, chemistry, electrochemical phenomena, and alloy evolution for molten salt equipment. In FY24, significant developments were made to the code package, enhancing its capabilities in many aspects. The improvements and advancements can be summarized as follows: 1. Implementation of tritium transport capabilities and validation with experimental data: To enable modeling of tritium and other fission gases within MSRs, we implemented gas transport within MOSCATO via inclusion of couple mass transport equations within the salt and structural alloys. Comparisons to experimental data from literature showed good agreement with respect to tritium release rates. 2. Preliminary implementation of two-phase flow models in MOSCATO: To model tritium and other gases above their solubility limits, we implemented preliminary two-phase flow models within MOSCATO to account for bubble transport. The first model adopted was the Level-Set approach, which can handle the high void fraction regime, but with a requirement for high mesh resolution thus high computational expense. In this report, we present a verification of the Level-Set method using a simple benchmark case. We also performed a demonstration of the code as applied to an experimental case involving cover gas flow through salt in an experimental vessel. The second model adopted was the Eulerian-Eulerian dispersed flow model, which is computationally cheaper but limited to low void fraction regimes, such as bubbly flow. Validation and verification have not yet been performed for the Eulerian-Eulerian approach, but a preliminary implementation was completed. 3. Validation with static corrosion experiments: Static corrosion experimental data for stainless steel coupons within molten salts was used to further validate the corrosion model in MOSCATO. To do so, we leveraged the existing models in MOSCATO and simulated the sample mass loss and mass gain phenomena. Several ion species, including Cr 2+ , Fe 2+ and H + , were simulated in salt using the PNP solver, while Cr 0 and Fe 0 were simulated with a diffusion solver in stainless steel. The mass loss of the samples was compared with experimental data, and good agreement was achieved. These combined activities served to further expand the capabilities of MOSCATO and make it more generally applicable to the full range of phenomena that can control chemistry and corrosion in molten salt reactors.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

DOE OSTI2021

Evaluating the Feasibility of Measuring and Analyzing Strain Caused by Changes in Barometric Pressure in the Vadose Zone

Changes in barometric pressure propagate into the subsurface where they can affect water level measurements in wells and cause vertical strain. Previous strain sensors were developed to measure vertical strain, but they were limited to measurements at a single depth, making it difficult to evaluate strains from a migrating pressure wave. Technology known as Coherence-length-gated Microwave Photonics Interferometry (CMPI) uses optical fiber sensors to measure strain at multiple locations. It has the potential to detect strain caused by variations in barometric pressure at multiple depths in the subsurface. The goal of this study is to evaluate the feasibility of using this technology to record the changes in vertical strain in the vadose zone from small fluctuations in air pressure. Propagation of these air pressure fluctuations is affected by hydrogeological properties such as water content and permeability. A second goal of this study is to evaluate how these properties affect the air pressure distribution, strain, and gas-phase diffusivity with a long-term goal of using strain to monitor the vadose zone. The focus of this study is on a suite of laboratory experiments that used a sand-filled column with an open head space. The CMPI fiber was packaged and embedded along the axis of the column to measure the strain at multiple locations while air pressure transducers were installed through the wall of the column. Small periodic fluctuations in air pressure, similar to barometric pressure fluctuations, were created in the head space of the column using an audio driven speaker driven at 4 Hz using a sinusoidal signal created by a function generator. Then, the fluctuations propagated along the column where they were measured by the transducers and recorded as functions of time. Initial tests were conducted using dry sand, but then additional tests were conducted after injecting water into the column, which changed the hydrologic and mechanical properties. Another experiment was conducted after creating a thin barrier to air flow, which was designed to simulate a thin layer of saturated ground during rainfall. A final experiment was conducted by changing the grain size of the material to create a heterogeneity in the upper half of the column. The suite of six experiments was conducted to highlight effects of variations in water content and heterogeneities on air pressure and vertical strain. Numerical simulations were also created to evaluate the laboratory data and to provide a baseline analysis for the results. The numerical simulations used the governing equations to linear poroelasticity and two-phase flow with boundary conditions representing the experiments. The conceptual model for strain caused by barometric pressure recognizes that a fluctuating barometric pressure causes fluctuations in the air pore pressure that decrease in amplitude and lag in time with increase depth. The pressure distribution results in two different loads that cause strain: 1.) a mechanical load as the barometric pressure acts on the ground surface; and 2.) pressure loading in the pore space. The results show that the total vertical strain is a contribution of both of these loads, and this is the strain that was observed in the laboratory. The lab experiments show that both air pressure and strain vary as sinusoids that propagate with depth. The amplitude of the air pressure and strain decrease as a function of depth in all the conditions that were evaluate. For example, in the dry sand, the amplitude of the air pressure decreases from 25 Pa at the head space to 5 Pa at a depth of 70 cm, whereas the amplitude of the strain decreases from 0.15 to 0.06 µε over the same depth interval. In many cases the amplitudes decrease is approximately negative exponential functions of depth. When water was injected into the sand and the saturation increased, the amplitude of the air pressure increased in the partially saturated sand but decreased sharply to zero where the soil was saturated. The strain decreased with depth, but the transition from partially to fully saturated conditions had little effect on the strain profile, even though it had a major effect on the air pressure. This apparently is because strain is caused by variations in both water and air pressure. The fluctuating air pressure caused the water pressure to fluctuate, which caused strain throughout the column. Phase delay of the air pressure and the strain both increase as linear to bi-linear functions of depth. The phase delays indicate that pressure propagates at a velocity of 7 to 8 m/s in dry or partially saturated sand, but it drops by more than an order of magnitude in the vicinity of the capillary fringe. The phase delays indicate the strain propagates faster than the pressure. The data indicate that the strain propagates from 1.2 to more than 4x faster than the pressure. Switching from sand to silt in the column caused the velocity of the pressure to drop by roughly half for both pressures, but curiously, it had little effect on the velocity of the strain.

54 ENVIRONMENTAL SCIENCES↗

DOE OSTI2026

Taylor approximation variance reduction for approximation errors in PDE-constrained Bayesian inverse problems

In numerous applications, surrogate models are used as a replacement for accurate parameter-to-observable mappings when solving large-scale inverse problems governed by partial differential equations (PDEs). The surrogate model may be a computationally cheaper alternative to the accurate parameter-to-observable mappings and/or may ignore additional unknowns or sources of uncertainty. The Bayesian approximation error (BAE) approach provides a means to account for the induced uncertainties and approximation errors, i.e. the errors between the accurate parameter-to-observable mapping and the surrogate. The statistics of these errors are, however, in general unknown a priori, and are thus calculated using Monte Carlo sampling. Although the sampling is typically carried out offline, i.e. before considering the data, the process can still represent a computational bottleneck. In this work, we develop a scalable computational approach for reducing the costs associated with the sampling stage of the BAE approach. Specifically, we consider the Taylor expansion of the accurate and surrogate forward models with respect to the uncertain parameter fields either as a control variate for variance reduction or as a means to directly and efficiently approximate the mean and covariance of the approximation errors. We propose efficient methods for evaluating the expressions for the mean and covariance of the Taylor approximations based on linear(-ized) PDE solves. Furthermore, the proposed approach is independent of the dimension of the uncertain parameter, depending instead on the intrinsic dimension of the data, ensuring scalability to high-dimensional problems. The potential benefits of the proposed approach are demonstrated for two high-dimensional inverse problems governed by PDE examples, namely for the estimation of a distributed Robin boundary coefficient in a linear diffusion problem, and for a coefficient estimation problem governed by a nonlinear diffusion problem.

Bayesian approximation error↗