SEARCH · Engineering Papers
Results for “spectral 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.
Towards a nonperturbative construction of the S-matrix
We present a nonperturbative recipe for directly computing the S-matrix in strongly-coupled QFTs. The method makes use of spectral data obtained in a Hamiltonian framework and can be applied to a wide range of theories, including potentially QCD. We demonstrate the utility of this prescription in the specific example of the 2+1d O(N) model at large N, using energy eigenstates computed with Hamiltonian truncation to reproduce the full 2 → 2 scattering amplitude for arbitrary (complex) center-of-mass energy.
Investigation of Cycle-to-Cycle Variations in Internal Combustion Engine Using Proper Orthogonal Decomposition
The understanding, modeling and control of the cycle-to-cycle variation (CCV) in the modern internal combustion engine (ICE) is a key scientific challenge to achieve stable engine operation. High CCV in the engine combustion chamber may contribute to partial burn, misfire and knock, which adversely affects the engine performance and may potentially damage the engine. The objective of the current study is to leverage high-fidelity numerical simulations to improve the understanding of the causes of CCV. Using the massively parallel code, Nek5000, multi-cycle, wall-resolved large-eddy simulations (LES) were performed for the General Motors (GM), Transparent Combustion Chamber (TCC-III) optical engine under motored operating conditions. Further, the large-scale structures of the in-cylinder flow were investigated using a triple proper orthogonal decomposition (POD) technique to explore the characteristics of different parts of the flow and their contributions to CCV. The kinetic energy of the subset of flow structures were determined and correlated between the intake and compression strokes. The insights from the analysis of the large-scale flow structures will be used to assist the development of improved engine designs with reduced CCV and enhance the engine performance.
Spectral induced polarization signatures of smoldering remediation enhanced with colloidal activated carbon: An experimental study
Monitoring the remediation of soil and groundwater contaminated by organic compounds remains highly challenging. The spectral induced polarization (SIP) method exhibits significant promise for monitoring changes to the electrical properties of soils that are undergoing remediation. Thermal treatments, such as smoldering combustion, have become established remediation techniques for destroying contaminants. Smoldering combustion is now being supported by colloidal activated carbon (CAC), with CAC able to adsorb contaminants and supplement the fuel source for destroying contaminants. The objective of this study is to investigate the potential of SIP for tracking the smoldering remediation of imitated field soils supplemented with CAC. SIP column experiments were first conducted to assess the response of SIP to varying concentrations of CAC in field soils that contain, or do not contain, organic material (OM). Here these results demonstrate that increasing OM and CAC contents increase both the real and imaginary components of the complex conductivity, with the imaginary conductivity also showing frequency dependence. Next, a suite of smoldering and SIP column experiments was conducted to investigate if SIP can detect changes in imitated field soils of varying OM and CAC contents that have been remediated by smoldering combustion. The SIP results on the examined soils both before and after smoldering show that SIP can detect changes in the real conductivity, and particularly the imaginary conductivity, between different soil compositions and different stages of the remediation process. High resolution scanning electron microscopy (SEM) imaging was performed on all samples to validate the SIP and smoldering experiments, confirming significant reductions in carbon after smoldering. Overall, this study suggests that SIP has potential to detect changes in the electrical properties of field soils due to the addition of remedial fluids like CAC and contaminant destruction by smoldering remediation.
High-order partitioned spectral deferred correction solvers for multiphysics problems
We present an arbitrarily high-order, conditionally stable, partitioned spectral deferred correction (SDC) method for solving multiphysics problems using a sequence of pre-existing single-physics solvers. This method extends the work in [1], [2], which used implicit-explicit Runge-Kutta methods (IMEX) to build high-order, partitioned multiphysics solvers. We consider a generic multiphysics problem modeled as a system of coupled ordinary differential equations (ODEs), coupled through coupling terms that can depend on the state of each subsystem; therefore the method applies to both a semi-discretized system of partial differential equations (PDEs) or problems naturally modeled as coupled systems of ODEs. The sufficient conditions to build arbitrarily high-order partitioned SDC schemes are derived. Based on these conditions, various of partitioned SDC schemes are designed. The stability of the first-order partitioned SDC scheme is analyzed in detail on a coupled, linear model problem. We show that the scheme is conditionally stable, and under conditions on the coupling strength, the scheme can be unconditionally stable. We demonstrate the performance of the proposed partitioned solvers on several classes of multiphysics problems with moderate coupling strength. They include a stiff linear system of ODEs, advection-diffusion-reaction systems, and fluid-structure interaction problems with both incompressible and compressible flows, where we verify the design order of the SDC schemes and study various stability properties. We also directly compare the accuracy, stability, and cost of the proposed partitioned SDC solver with the partitioned IMEX method in [1], [2] on this suite of test problems. The results suggest that the high-order partitioned SDC solvers are more robust than the partitioned IMEX solvers for the numerical examples considered in this work, while the IMEX methods require fewer implicit solves.
NSFnets (Navier-Stokes flow nets): Physics-informed neural networks for the incompressible Navier-Stokes equations
In the last 50 years there has been a tremendous progress in solving numerically the Navier-Stokes equations using finite differences, finite elements, spectral, and even meshless methods. Yet, in many real cases, we still cannot incorporate seamlessly (multi-fidelity) data into existing algorithms, and for industrial-complexity applications the mesh generation is time consuming and still an art. Moreover, solving ill-posed problems (e.g., lacking boundary conditions) or inverse problems is often prohibitively expensive and requires different formulations and new computer codes. Here, we employ physics-informed neural networks (PINNs), encoding the governing equations directly into the deep neural network via automatic differentiation, to overcome some of the aforementioned limitations for simulating incompressible laminar and turbulent flows. We develop the Navier-Stokes flow nets (NSFnets) by considering two different mathematical formulations of the Navier-Stokes equations: the velocity-pressure (VP) formulation and the vorticity-velocity (VV) formulation. Since this is a new approach, we first select some standard benchmark problems to assess the accuracy, convergence rate, computational cost and flexibility of NSFnets; analytical solutions and direct numerical simulation (DNS) databases provide proper initial and boundary conditions for the NSFnet simulations. The spatial and temporal coordinates are the inputs of the NSFnets, while the instantaneous velocity and pressure fields are the outputs for the VP-NSFnet, and the instantaneous velocity and vorticity fields are the outputs for the VV-NSFnet. This is unsupervised learning and, hence, no labeled data are required beyond boundary and initial conditions and the fluid properties. The residuals of the VP or VV governing equations, together with the initial and boundary conditions, are embedded into the loss function of the NSFnets. No data is provided for the pressure to the VP-NSFnet, which is a hidden state and is obtained via the incompressibility constraint without extra computational cost. Unlike the traditional numerical methods, NSFnets inherit the properties of neural networks (NNs), hence the total error is composed of the approximation, the optimization, and the generalization errors. Here, we empirically attempt to quantify these errors by varying the sampling (“residual”) points, the iterative solvers, and the size of the NN architecture. For the laminar flow solutions, we show that both the VP and the VV formulations are comparable in accuracy but their best performance corresponds to different NN architectures. The initial convergence rate is fast but the error eventually saturates to a plateau due to the dominance of the optimization error. For the turbulent channel flow, we show that NSFnets can sustain turbulence at , but due to expensive training we only consider part of the channel domain and enforce velocity boundary conditions on the subdomain boundaries provided by the DNS data base. We also perform a systematic study on the weights used in the loss function for balancing the data and physics components, and investigate a new way of computing the weights dynamically to accelerate training and enhance accuracy. In the last part, we demonstrate how NSFnets should be used in practice, namely for ill-posed problems with incomplete or noisy boundary conditions as well as for inverse problems. We obtain reasonably accurate solutions for such cases as well without the need to change the NSFnets and at the same computational cost as in the forward well-posed problems. As a result, we also present a simple example of transfer learning that will aid in accelerating the training of NSFnets for different parameter settings.
Dynamics of reconfigurable artificial spin ice: Toward magnonic functional materials
Over the past few years, the study of magnetization dynamics in artificial spin ices has become a vibrant field of study. Artificial spin ices are ensembles of geometrically arranged, interacting magnetic nanoislands, which display frustration by design. These were initially created to mimic the behavior in rare earth pyrochlore materials and to study emergent behavior and frustration using two-dimensional magnetic measurement techniques. Recently, it has become clear that it is possible to create artificial spin ices, which can potentially be used as functional materials. In this perspective, we review the resonant behavior of spin ices in the GHz frequency range, focusing on their potential application as magnonic crystals. In magnonic crystals, spin waves are functionalized for logic applications by means of band structure engineering. While it has been established that artificial spin ices can possess rich mode spectra, the applicability of spin ices to create magnonic crystals hinges upon their reconfigurability. Consequently, we describe recent work aiming to develop techniques and create geometries allowing full reconfigurability of the spin ice magnetic state. We also discuss experimental, theoretical, and numerical methods for determining the spectral response of artificial spin ices and give an outlook on new directions for reconfigurable spin ices.
Two-dimensional electronic–vibrational spectroscopy: Exploring the interplay of electrons and nuclei in excited state molecular dynamics
Two-dimensional electronic-vibrational spectroscopy (2DEVS) is an emerging spectroscopic technique which exploits two different frequency ranges for the excitation (visible) and detection (infrared) axes of a 2D spectrum. In contrast to degenerate 2D techniques, such as 2D electronic or 2D infrared spectroscopy, the spectral features of a 2DEV spectrum report cross correlations between fluctuating electronic and vibrational energy gaps rather than autocorrelations as in the degenerate spectroscopies. The center line slope of the spectral features reports on this cross correlation function directly and can reveal specific electronic-vibrational couplings and rapid changes in the electronic structure, for example. The involvement of the two types of transition moments, visible and infrared, makes 2DEVS very sensitive to electronic and vibronic mixing. 2DEV spectra also feature improved spectral resolution, making the method valuable for unraveling the highly congested spectra of molecular complexes. The unique features of 2DEVS are illustrated in this paper with specific examples and their origin described at an intuitive level with references to formal derivations provided. Although early in its development and far from fully explored, 2DEVS has already proven to be a valuable addition to the tool box of ultrafast nonlinear optical spectroscopy and is of promising potential in future efforts to explore the intricate connection between electronic and vibrational nuclear degrees of freedom in energy and charge transport applications.
The PAU Survey: an improved photo- z sample in the COSMOS field
Here we present – and make publicly available – accurate and precise photometric redshifts in the ACS footprint from the COSMOS field for objects with i AB ≤ 23. The redshifts are computed using a combination of narrow-band photometry from PAUS, a survey with 40 narrow bands spaced at $100\,\mathring{\rm A}$ intervals covering the range from 4500 to $8500\,\mathring{\rm A}$, and 26 broad, intermediate, and narrow bands covering the UV, visible and near-infrared spectrum from the COSMOS2015 catalogue. We introduce a new method that models the spectral energy distributions as a linear combination of continuum and emission-line templates and computes its Bayes evidence, integrating over the linear combinations. The correlation between the UV luminosity and the $\mathrm{O\,{\small II}}$ line is measured using the 66 available bands with the zCOSMOS spectroscopic sample, and used as a prior which constrains the relative flux between continuum and emission-line templates. The flux ratios between the $\mathrm{O\,{\small II}}$ line and H α , H β and $\mathrm{O\,{\small III}}$ are similarly measured and used to generate the emission-line templates. Comparing to public spectroscopic surveys via the quantity Δ z ≡ (z photo – z spec )/(1 + z spec ), we find the photometric redshifts to be more precise than previous estimates, with σ 68 (Δ z ) ≈ (0.003, 0.009) for galaxies at magnitude i AB ~ 18 and i AB ~ 23, respectively, which is three times and 1.66 times tighter than COSMOS2015. Additionally, we find the redshifts to be very accurate on average, yielding a median of the Δ z distribution compatible with |median(Δ z )| ≤ 0.001 at all redshifts and magnitudes considered. Both the added PAUS data and new methodology contribute significantly to the improved results. The catalogue produced with the technique presented here is expected to provide a robust redshift calibration for current and future lensing surveys, and allows one to probe galaxy formation physics in an unexplored luminosity-redshift regime, thanks to its combination of depth, completeness, and excellent redshift precision and accuracy.
Investigating the Origins of Cyclic Variability in Internal Combustion Engines Using Wall-Resolved Large Eddy Simulations
Modern internal combustion engines (ICE) operate at the ragged edge of stable operation characterized by high cycle-to-cycle variations (CCV). A key scientific challenge for ICE is the understanding, modeling, and control of CCV in engine performance, which can contribute to partial burns, misfire, and knock. The objective of this study is to use high-fidelity numerical simulations to improve the understanding of the causes of CCV. Nek5000, a leading high-order spectral element, open source code, is used to simulate the turbulent flow in the engine combustion chamber. Multicycle, wall-resolved large-eddy simulations (LESs) are performed for the General Motors (GM), Transparent Combustion Chamber (TCC-III) optical engine under motored operating conditions. The mean and root-mean-square (rms) of the in-cylinder flow fields at various piston positions are validated using particle image velocimetry (PIV) measurements during the intake and compression strokes. The large-scale flow structures, including the swirl and tumble flow patterns, are analyzed in detail and the causes for cyclic variabilities in these flow features are explained. The energy distribution across the different scales of the flow are quantified using one-dimensional (1D) energy spectra, and the effect of the tumble breakdown process on the energy distribution is examined. Finally, the insights from this study can help us develop improved engine designs with reduced cyclic variabilities in the in-cylinder flow leading to enhanced engine performance.
Significant DBSCAN+: Statistically Robust Density-based Clustering
Cluster detection is important and widely used in a variety of applications, including public health, public safety, transportation, and so on. Given a collection of data points, we aim to detect density-connected spatial clusters with varying geometric shapes and densities, under the constraint that the clusters are statistically significant. The problem is challenging, because many societal applications and domain science studies have low tolerance for spurious results, and clusters may have arbitrary shapes and varying densities. As a classical topic in data mining and learning, a myriad of techniques have been developed to detect clusters with both varying shapes and densities (e.g., density-based, hierarchical, spectral, or deep clustering methods). However, the vast majority of these techniques do not consider statistical rigor and are susceptible to detecting spurious clusters formed as a result of natural randomness. On the other hand, scan statistic approaches explicitly control the rate of spurious results, but they typically assume a single “hotspot” of over-density and many rely on further assumptions such as a tessellated input space. To unite the strengths of both lines of work, we propose a statistically robust formulation of a multi-scale DBSCAN, namely Significant DBSCAN+, to identify significant clusters that are density connected. As we will show, incorporation of statistical rigor is a powerful mechanism that allows the new Significant DBSCAN+ to outperform state-of-the-art clustering techniques in various scenarios. We also propose computational enhancements to speed-up the proposed approach. Experiment results show that Significant DBSCAN+ can simultaneously improve the success rate of true cluster detection (e.g., 10–20% increases in absolute F1 scores) and substantially reduce the rate of spurious results (e.g., from thousands/hundreds of spurious detections to none or just a few across 100 datasets), and the acceleration methods can improve the efficiency for both clustered and non-clustered data.
Implementation of stacked ensemble machine learning for the detection of surrogate plutonium contamination in soil via LIBS
Supervised machine learning methods have demonstrated increased utility for the quantification of lanthanide and actinide elements in atomic spectroscopy applications. This study implements laser-induced breakdown spectroscopy (LIBS) for the identification of plutonium surrogate material (CeO 2 ) in soil matrices by training supervised machine learning methods on the recorded spectral data. A bagged ensemble using Random Forest yields the highest sensitivity predictions with a detection limit of 0.015 wt.% CeO 2 . However, high precision in Ce content prediction required the use of a stacked ensemble regression, which provided the superlative Ce quantification model with an error of 0.107% and a detection limit of 0.022 wt.%. Furthermore, the high performance of the stacked ensemble demonstrates its potential to enhance the accuracy and sensitivity of nuclear contaminant detection using field-deployable spectroscopic analyzers in real-world scenarios.
2021 Summer SPE Project
SPE is a project in order to develop new, more physics-based, seismic models of explosions (see Nelson et al. 2013). One key component of this effort is numerical modeling enabled by modern state-of-art code/software. Accurate modeling of the shape and amplitude of seismic waves from their generation to their propagation to remote monitoring seismic stations is important to our ability to determine the origin and strength of the source from remote recording. In this project, the modeling is performed by coupling two codes HOSS and SPECFEM3D. HOSS models the dynamic nonlinear processes happening near the explosion. SPECFEM3D computes the propagation of seismic waves as they travel through 3D complex Earth models. SPECFEM3D was modified in order to be driven by a set of time-series calculated by HOSS in lieu of a seismic source. The goal of this summer project is the investigation of several questions pertinent to the establishment of a full end-to-end modeling capability from the high-rate strain regime area to remote distances where seismic station record seismic waves generated by explosions. The investigated questions are: (1) How to perform proper filtering? Direct-solution modeling only sustains a limited range of frequency depending on the grid size. As we go from one modeling domain to the other via coupling, the mesh size is getting coarser to allow modeling at large scale but also to account for the fact that high-frequency waves do not physically travel to large distances. So filtering of the time-series generated by the near-field hydrodynamic modeling is a current practice often employed but its effects on the modeled waveforms has to be investigated. (2) Quantitative assessment of the efficiency of attenuation to remove high-frequency content of the wavefield. This assessment will allow to create meshes with a grid size appropriate to the actual physics of wave propagation for a given explosion. (3) Checking that the coupling process and the two codes respect the cylindrical symmetry that is expected in the case of a pure explosion in a half-space. (4) The effect of the state-of-stress in the near-source area on the modeled seismic waveforms. These questions will be investigated through the modeling of SPE-4P, the fourth explosion of this series because it was designed to have little interaction with the geologic surrounding and the free surface of the Earth so that it is the most explosion-like experiment, with the most symmetries to be verified.
Scalable Interpolation on GPUs for Thermal Fluids Applications
The need for efficient interpolation arises in many use cases for spectral element simulations. Towards this end, we present a fast and robust GPU-enable interpolation utility for NekRS, a spectral-element, Navier-Stokes solver. This utility is based on the findpts and findpts eval routines from the gslib library. To demonstrate the functionality, we added support for particle tracking. The GPU implementation of the interpolation allows for tracking 100 times more particles for the same computational cost.
High-Order Hybrid RANS-LES Study of NACA0012 Wing Sections
We develop hybrid RANS-LES strategies within Nek5000 for application to airfoil sections at small flight configurations. We present a validation and verification study of $k \space – \space \tau$ SST applied to a NACA 0012 wing section in a pure RANS and in a hybrid RANS-LES setup. The study shows good corroboration with existing experimental and numerical datasets. We also analyze some of the observed discrepancies with the experiments by evaluating the side wall “blocking” effect. We demonstrate that for the hybrid turbulence modeling approach a high-order spectral- element discretization converges faster (i.e., with less resolution) than a representative low-order finite-volume-based approach.
The K2 M67 Study: Precise Mass for a Turnoff Star in the Old Open Cluster M67
We present a study of the bright detached eclipsing main-sequence binary WOCS 11028 (Sanders 617) in the open cluster M67. Although the binary has only one eclipse per orbital cycle, we show that the masses of the stars can be derived very precisely thanks to a strong constraint on the orbital inclination: M {sub A} = 1.222 ± 0.006M {sub ⊙} and M {sub B} = 0.909 ± 0.004M {sub ⊙}. We use a spectral energy distribution fitting method to derive characteristics of the component stars in lieu of the precise radii that would normally be derived from a doubly eclipsing binary. The deconvolution of the SEDs reveals that the brighter component of the binary is at the faint turnoff point for the cluster—a distinct evolutionary point that occurs after the convective core has been established and while the star is in the middle of its movement toward lower surface temperatures, before the so-called hook at the end of the main sequence. The measurements are in distinct disagreement with evolution models at solar metallicity: higher metal abundances are needed to reproduce the characteristics of WOCS 11028 A. We discuss the changes to model physics that are likely to be needed to address the discrepancies. The clearest conclusions are that diffusion is probably necessary to reconcile spectroscopic abundances of M67 stars with the need for higher metallicity models and that reduced strength convective overshooting is occurring for stars at the turnoff. At super-solar bulk metallicity, various indicators agree on a cluster age between about 3.5 and 4.0 Gyr.
Reduced-Order Models of Static Power Grids based on Spectral Clustering
For large-scale interconnected power systems that cover large geographical areas, certain electrical studies are required so that appropriate decisions ensure system reliability and low cost. For such studies, it is often neither practical nor necessary to model in detail the entire power system, which is increasingly complex due to a more diverse range of grid assets to choose from in both short and long-term planning. The goal of this paper is to present a methodology to reduce the order of large-scale power networks based on spectral graph theory given that current methods for static network reduction are not scalable. A brief analysis of some spectral clustering properties to determine which graph Laplacian matrix should be used and why is included. The analysis shows that the utilization of the normalized graph Laplacian is more advantageous for clustering purposes. Techniques are proposed to approximate cost functions for the aggregated generators. This is done via linear regression. The reduced-order model obtained with the proposed methodology has an accuracy above 94% and solves the scalability issue commonly present in other reduction methods. If the utilization of the reduced-order model is either constrained to load levels above mid-peak demand, or cost functions of aggregated units are approximated via a piecewise quadratic approach, then the error distribution is in the order of 10^-3. .
Characterizing Physical Properties of Streambed Interface Sediments Using In Situ Complex Electrical Conductivity Measurements
Streambed sediment physical properties such as surface area, are difficult to quantify in situ but exert a high-level control on a wide range of biogeochemical processes and sorption of contaminants. We introduce the use of complex electrical conductivity (CC) methods (also known as spectral-induced polarization (SIP)) that measure both real and imaginary conductivity to noninvasively and efficiently characterize shallow streambed sediments. We explore the method through synthetic modeling, laboratory, and field measurements to demonstrate the sensitivity of imaginary conductivity to sediment surface area, controlled in part by fine-grained iron oxides produced by anoxic groundwater discharge. Laboratory measurements verify expected relationships between CC parameters and sediment properties. Synthetic modeling using a 1D analytical model illustrates the influence of water layer depth and conductivity on the field CC measurements made at the streambed-stream water interface. Specifically, the inverted sediment imaginary conductivity is less impacted by uncertainty in the water layer depth and conductivity relative to the real conductivity and phase shift. Field CC measurements along a landfill-impacted river reveal discrete streambed zones with enhanced bulk surface area generally corresponding to anoxic groundwater discharge zones with high concentrations of fine-grained iron oxide precipitates.