SEARCH · Search NASA

Results for “Coupled process modeling”

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 1,729 records · Page 96

All sources NASA NTRS DOE OSTI

NASA NTRS2022

Updraft Dynamics and Microphysics: On the Added Value of The Cumulus Thermal Reference Frame in Simulations of Aerosol–Deep Convection Interactions

One fundamental question about atmospheric moist convection processes that remains debated is whether, or under which conditions, a relevant variability in background aerosol concentrations may have a significant dynamical impact on convective clouds and their associated precipitation. Furthermore, current climate models must parameterize both the microphysical and the cumulus convection processes, but this is usually implemented separately, whereas in nature there is a strong coupling between them. As a first step to improve our understanding of these two problems, we investigate how aerosol concentrations modify key properties of updrafts in eight large-eddy-permitting regional simulations of a case study of scattered convection over Houston, Texas, in which convection is explicitly simulated and microphysical processes are parameterized. Dynamical and liquid-phase microphysical responses are investigated using the following two different reference frames: static cloudy updraft grid cells versus tracked cumulus thermals. In both frameworks, we observe the expected microphysical responses to higher aerosol concentrations, such as higher cloud number concentrations and lower rain number concentrations. In terms of the dynamical responses, both frameworks indicate weak impacts of varying aerosol concentrations relative to the noise between simulations over the observationally derived range of aerosol variability for this case study. On the other hand, results suggest that thermals are more selective than cloudy updraft grid cells in terms of sampling the most active convective air masses. For instance, vertical velocity from thermals is significantly higher at upper levels than when sampled from cloudy updraft grid points, and several microphysical variables have higher average values in the cumulus thermal framework than in the cloudy updraft framework. In addition, the thermal analysis is seen to add rich quantitative information about the rates and covariability of microphysical processes spatially and throughout tracked thermal lifecycles, which can serve as a stronger foundation for improving subgrid-scale parameterizations.

aerosols↗

NASA NTRS2022

Updraft Dynamics and Microphysics: On the Added Value of the Cumulus Thermal Reference Frame in Simulations of Aerosol–Deep Convection Interactions

One fundamental question about atmospheric moist convection processes that remains debated is whether or under what conditions a relevant variability in background aerosol concentrations may have a significant dynamical impact on convective clouds and their associated precipitation. Furthermore, current climate models must parameterize both the microphysical and the cumulus convection processes, but this is usually implemented separately, whereas in nature there is a strong coupling between them. As a first step to improve our understanding of these two problems, we investigate how aerosol concentrations modify key properties of updrafts in eight large-eddy permitting regional simulations of a case study of scattered convection over Houston, Texas, in which convection is explicitly simulated and microphysical processes are parameterized. Dynamical and liquid-phase microphysical responses are investigated using two different reference frames: static cloudy-updraft grid cells versus tracked cumulus thermals. In both frameworks we observe the expected microphysical responses to higher aerosol concentrations, such as higher cloud number concentrations and lower rain number concentrations. In terms of the dynamical responses, both frameworks indicate weak impacts of varying aerosol concentrations relative to the noise between simulations over the observationally derived range of aerosol variability for this case study. On the other hand, results suggest that thermals are more selective than cloudy-updraft grid cells in terms of sampling the most active convective air masses. For instance, vertical velocity from thermals is significantly higher at upper levels than when sampled from cloudy-updraft grid points, and several microphysical variables have higher average values in the cumulus thermal framework than in the cloudy-updraft framework. In addition, the thermal analysis is seen to add rich quantitative information about the rates and covariability of microphysical processes spatially and throughout tracked thermal lifecycles, which can serve as a stronger foundation for improving subgrid-scale parameterizations.

Daniel Hernandez-Deckers↗

DOE OSTI2012

Two- and Three-Body Charmless B Decays at BaBar

We report recent measurements of rare charmless B decays performed by BaBar. The results are based on the final BaBar dataset of 424 fb{sup -1} collected at the PEP-II B-factory based at the SLAC National Accelerator Laboratory. The study of rare B decays is a key ingredient to meet two of the main goals of the B-factories: assessing the validity of the Cabibbo-Kobayashi-Maskawa (CKM) picture of CP-violation by precisely measuring the elements of the Unitarity Triangle (UT), and searching for hints of New Physics (NP), or otherwise constraining NP scenarios, in processes which are suppressed in the Standard Model (SM). In loop processes, in particular, NP at some higher energy scale may manifest itself in the low energy effective theory as new couplings, such as those introduced by new very massive virtual particles in the loop. In NP searches hadronic uncertainties can play a major role, expecially for branching fraction measurements. Many theoretical uncertainties cancel in ratios of amplitudes, and most NP probes are therefore of this kind. In the following sections we report recent measurements, performed by the BaBar Collaboration, that are relevant to NP searches in charmless hadronic B decays.

Experiment-HEP,HEPEX↗

DOE OSTI2011

Higgs quadruplet for the type III seesaw model and implications for {mu}{yields}e{gamma} and {mu}-e conversion

In the type III seesaw model, the heavy neutrinos are contained in leptonic triplet representations. The Yukawa couplings of the triplet fermion and the left-handed neutrinos with the doublet Higgs field produce the Dirac mass terms. Together with the Majorana masses for the leptonic triplets, the light neutrinos obtain nonzero seesaw masses. We point out that it is also possible to have a quadruplet Higgs field to produce the Dirac mass terms to facilitate the seesaw mechanism. The vacuum expectation value of the quadruplet Higgs is constrained to be small by electroweak precision data. Therefore, the Yukawa couplings of a quadruplet can be much larger than those for a doublet. We also find that unlike the usual type III seesaw model where at least two copies of leptonic triplets are needed, with both doublet and quadruplet Higgs representations, just one leptonic triplet is possible to have a phenomenologically acceptable model because light neutrino masses can receive sizable contributions at both tree and one-loop levels. Large Yukawa couplings of the quadruplet can induce observable effects for lepton flavor violating processes {mu}{yields}e{gamma} and {mu}-e conversion. Implications of the recent {mu}{yields}e{gamma} limit from MEG and the limit on {mu}-e conversion on Au are also given. Some interesting collider signatures for the doubly charged Higgs boson in the quadruplet are discussed.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

DOE OSTI2025

A theoretical model for sheath dynamics in the two-plasma mode of a cylindrical filament discharge

We present a refined understanding of cylindrical plasma-facing emitter cathodes by examining the limitations of the conventional space charge limited model, which overlooks ion trapping dynamics within the virtual cathode. While conventional theory distinguishes between temperature-limited and space charge limited regimes based on electron current constraints, our findings reveal that trapped ions can form a quasi-neutral “second plasma,” expanding upstream and interacting with the primary plasma. In cylindrical and other non-planar geometries, an expanding trapped-ions layer produces current enhancement by converting more of the emitted electron flux to passing flux. The current enhancement leads to a complex coupling between the trapped-ions plasma and upstream plasma through ionizing collisions in the upstream region and charge exchange collisions in the virtual cathode. The coupling results in mutual density enhancement and at the same time, antagonistic expansion dynamics between the two plasma layers. The process is understood using a new “aid-compete” model. Numerical solutions, validated against simulations, reveal the model's capacity to predict system evolution and complex discharge behaviors, albeit with qualitative approximations.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

DOE OSTI2016

New insights into the nonadiabatic state population dynamics of model proton-coupled electron transfer reactions from the mixed quantum-classical Liouville approach

In a previous study [F. A. Shakib and G. Hanna, J. Chem. Phys. 141, 044122 (2014)], we investigated a model proton-coupled electron transfer (PCET) reaction via the mixed quantum-classical Liouville (MQCL) approach and found that the trajectories spend the majority of their time on the mean of two coherently coupled adiabatic potential energy surfaces. This suggested a need for mean surface evolution to accurately simulate observables related to ultrafast PCET processes. In this study, we simulate the time-dependent populations of the three lowest adiabatic states in the ET-PT (i.e., electron transfer preceding proton transfer) version of the same PCET model via the MQCL approach and compare them to the exact quantum results and those obtained via the fewest switches surface hopping (FSSH) approach. We find that the MQCL population profiles are in good agreement with the exact quantum results and show a significant improvement over the FSSH results. All of the mean surfaces are shown to play a direct role in the dynamics of the state populations. Interestingly, our results indicate that the population transfer to the second-excited state can be mediated by dynamics on the mean of the ground and second-excited state surfaces, as part of a sequence of nonadiabatic transitions that bypasses the first-excited state surface altogether. This is made possible through nonadiabatic transitions between different mean surfaces, which is the manifestation of coherence transfer in MQCL dynamics. We also investigate the effect of the strength of the coupling between the proton/electron and the solvent coordinate on the state population dynamics. Drastic changes in the population dynamics are observed, which can be understood in terms of the changes in the potential energy surfaces and the nonadiabatic couplings. Finally, we investigate the state population dynamics in the PT-ET (i.e., proton transfer preceding electron transfer) and concerted versions of the model. The PT-ET results confirm the participation of all of the mean surfaces, albeit in different proportions compared to the ET-PT case, while the concerted results indicate that the mean of the ground- and first-excited state surfaces only plays a role, due to the large energy gaps between the ground- and second-excited state surfaces.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

NASA NTRS2020

Desorption Kinetics of O and CO from Graphitic Carbon Surfaces

The desorption of O/CO from graphitic carbon surfaces is investigated using a one-dimensional model describing the adsorbate interactions with the surface phonon bath. The kinetics of desorption are described through the solution of a master equation for the time-dependent population of the adsorbate in an oscillator state, which is modified through thermal fluctuations at the surface. The interaction of the adsorbate with the surface phonons is explicitly captured by using the computed phonon density Of states (PDOS) of the surface. The coupling of the adsorbate with the phonon bath results in the transition of the adsorbate up and down a vibrational ladder. The adsorbate-surface interaction is represented in the model using a Morse potential, which allows for the desorption process to be directly modeled as a transition from bound to free (continuum) state. The PDOS is a property of the material and the lattice; and is highly sensitive to the presence of defects. The effect of etch pits along with random surface defects on the PDOS is considered in the present work. The presence of defects causes a redshift and broadening of the PDOS, which in turn changes the phonon frequency modes available for adsorbate coupling at the surface. Using the realistic PDOS distributions, the phonon-induced desorption (PID) model was used to compute the transition and desorption rates for both pristine and defective systems. Mathissen’s rule is used to compute the phonon relaxation time for pristine and defective systems based on the phonon scattering times for each of the different scattering processes. First, the desorption rates of the pristine system is fitted against the experimental values to obtain the Morse potential parameters for each of the observed adatoms. These Morse potential parameters are used along with the defective PDOS and phonon relaxation time to compute the desorption rates for the defective system. The defective system rates (both transition and desorption) were consistently lower in comparison with the pristine system. The difference between the transition rates is more significant at lower initial states due to higher energy spacing between the levels. In the case of the desorption rates, the difference between the defective and pristine system is more significant at higher temperatures. The desorption rates for each of the system shows an order of magnitude decrease with the strongly bound systems exhibiting the greatest reduction in the desorption rates.

Swaminathan-Gopalan, Krishnan↗

DOE OSTI2023

High-Fidelity Energy Deposition Ignition Model Coupled with Flame Propagation Models at Engine-like Flow Conditions

With the heightened pressure on car manufacturers to increase the efficiency and reduce the carbon emissions of their fleets, more challenging engine operation has become a viable option. Highly dilute, boosted, and stratified charge, among others, promise engine efficiency gains and emissions reductions. At such demanding engine conditions, the spark-ignition process is a key factor for the flame initiation propagation and the combustion event. From a computational standpoint, there exist multiple spark-ignition models that perform well under conventional conditions but are not truly predictive under strenuous engine operation modes, where the underlying physics needs to be expanded. In this paper, a hybrid Lagrangian-Eulerian spark-ignition (LESI) model is coupled with different turbulence models, grid sizes, and combustion models. The ignition model, previously developed, relies on coupling Eulerian energy deposition with a Lagrangian particle evolution of the spark channel, at every time-step. The spark channel is attached to the electrodes and allowed to elongate at a speed derived from the flow velocity. The LESI model is used to simulate spark ignition in a nonquiescent crossflow environment at engine-like conditions, using converge commercial computational fluid dynamics (CFD) solver. The results highlight the consistency, robustness, and versatility of the model in a range of engine-like setups, from typical with Reynolds-averaged Navier-Stokes (RANS) and a larger grid size to high fidelity with large-eddy simulation (LES) and a finer grid size. The flame kernel growth is then evaluated against Schlieren images from an optical constant volume ignition chamber with a focus on the performance of flame propagation models, such as G-equation and thickened flame model, versus the baseline well-stirred reactor model. Finally, future development details are discussed.

Advanced ignition modeling↗

NASA NTRS2024

Replicating the Hadley Cell Edge and Subtropical Jet Latitude Disconnect in Idealized Atmospheric Models

Recent work has shown that variability in the subtropical jet's (STJ) latitude, ϕSTJ, is not coupled to that of the Hadley cell (HC) edge, ϕHC, but the robustness of this disconnect has not been examined in detail. Here, we use meteorological reanalysis products, comprehensive climate models, and an idealized atmospheric model to determine the necessary processes for a disconnect between ϕHC and ϕSTJ in the Northern Hemisphere's December–January–February season. We find that a decoupling can occur in a dry general circulation model, indicating that large-scale dynamical processes are sufficient to reproduce the metrics' relationship. It is therefore not reliant on explicit variability in the zonal structure, convection, or radiation. Rather, the disconnect requires a sufficiently realistic climatological basic state. Further, we confirm that the robust disconnect between ϕSTJ and ϕHC across the model hierarchy reveals their differing sensitivities to midlatitude eddy momentum fluxes; ϕHC is consistently coupled to the latitude of maximum eddy momentum flux, but ϕSTJ is not.

Molly E. Menzel↗

NASA NTRS2003

Flow Field Measurement of Mixing Driven by Buoyancy

Mixing driven by buoyancy-induced flows inside a cavity consists of stretching and folding of an interface. Measurement of the flow field using particle imaging velocimetry shows that during stretching the flow field has a single elliptic point, thus dominated by a single vortex. However, global bifurcation that results in folding introduces a hyperbolic point whereby the flow field degenerates to multiple vortex interactions. The short-lived coherent structure observed during mixing which results in the Rayleigh- Taylor morphology is attributed to vortex interactions. The mixing characteristics of non-homogeneous fluids driven by buoyancy are important towards understanding transport phenomenon in a microgravity environment. Mixing consists of stretching and folding of an interface due to a flow field whose intensity depends on the body force. For miscible liquids, the characteristic of the flow field determines whether mass transport is governed by diffusion or bulk stirring which induces mixing. For technologically important processes, transport of mass is governed by the coupling of the body force to scalar gradients such as concentration and or temperature' 2 3 . In order to lend insight into these classes of problems we consider a model experimental system to study mixing driven by buoyancy-induced flows. The characteristics of mixing is addressed from detail measurements of the flow field using particle imaging velocimetry (PIV), and its corresponding interface dynamics using image processing techniques.

Batur, C.↗

DOE OSTI2022

Report of the Frontier for Rare Processes and Precision Measurements

This is the Snowmass 2021 Rare and Precision Frontier Report. The Rare Processes and Precision Measurements Frontier, referred to as the ``Rare and Precision Frontier", or RPF, encompasses searches for extremely rare processes or tiny deviations from the Standard Model (SM) that can be studied with intense sources and high-precision detectors. Our community studies have identified several unique research opportunities that may pin down the scales associated with New Physics (NP) interactions and constrain the couplings of possible new degrees of freedom. Searches for rare flavor transition processes and precision measurements are indispensable probes of flavor and fundamental symmetries, and provide insights into physics that manifests itself at higher energy or through weaker interactions than those directly accessible at high-energy colliders.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

DOE OSTI2023

Microstructure-process relationships in monolithic U-10Mo fuel foil single-pass rolling: A parametric simulation study

In this work, a previously validated coupling of Kinetic Monte Carlo (KMC) Potts Model and finite element method (FEM) simulations was implemented to investigate the effects of microstructural features in as-cast and homogenized monolithic U-10Mo foils on the emergent microstructure after rolling and reheating. Parameters that could potentially affect recrystallization behavior of the rolled U-10Mo foils were considered: grain size distribution, uranium carbide (UC) size distribution, UC volume fraction, spatial distribution of UC, and rolling reduction magnitude. Grain structure and the magnitude of rolling reduction have the strongest influence on recrystallization kinetics and the fabricated grain size distribution. The UC distribution had only a weak effect on the recrystallization kinetics and final microstructures. While particle-stimulated nucleation (PSN) occurred in simulation more frequently as grain size increased, its incidence did not appear to considerably affect the recrystallization kinetics or grain size distribution.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

NASA NTRS1991

Numerical investigations with a hybrid isentropic-sigma model. I - Normal-mode characteristics. II - The inclusion of moist processes

The normal-mode characteristics of baroclinically amplifying disturbances were numerically investigated in a series of adiabatic simulations by a hybrid isentropic-sigma model, demonstrating the effect of coupling an isentropic-coordinate free atmospheric domain with a sigma-coordinate PBL on the normal-mode characteristics. Next, the normal-mode model was modified by including a transport equation for water vapor and adiabatic heating by condensation. Simulations with and without a hydrological component showed that the overall effect of latent heat release is to markedly enhance cyclogenesis and frontogenesis.

Pierce, R. B.↗

DOE OSTI2017

Soot and Spectral Radiation Modeling in ECN Spray A and in Engines

The amount of soot formed in a turbulent combustion system is determined by a complex system of coupled nonlinear chemical and physical processes. Different physical subprocesses can dominate, depending on the hydrodynamic and thermochemical environments. Similarly, the relative importance of reabsorption, spectral radiation properties, and molecular gas radiation versus soot radiation varies with thermochemical conditions, and in ways that are difficult to predict for the highly nonhomogeneous in-cylinder mixtures in engines. Here it is shown that transport and mixing play relatively more important roles as rate-determining processes in soot formation at engine-relevant conditions. It is also shown that molecular gas radiation and spectral radiation properties are important for engine-relevant conditions.

33 ADVANCED PROPULSION SYSTEMS↗

DOE OSTI2018

Argo Observations of the Deep Mixing Band in the Southern Ocean: A Salinity Modeling Challenge

The Southern Ocean plays an important role in mediating oceanic uptake of CO 2 and heat due to a strong meridional overturning circulation. Gridded Argo float data for 2004–2017 were used to evaluate subsurface processes at the mixed layer depth (MLD) that occur in a narrow deep mixing band. Shifts in the value of the Turner Angle at the MLD indicate that early in the season the MLD deepens slowly as it encounters and is stabilized by a subsurface salt maximum. By September mixing has penetrated this salinity feature and the rate of deepening is faster once the MLD is deeper than the depth where the maximum salinity occurs (~150–200 m). This distinctive salinity layer is the result of surface Ekman transport of fresh water from the south and subsurface advection of high-salinity water from the north. Two configurations of the Community Earth System Model (CESM) ocean-ice forced hindcast experiments—one with 1° and the other with 0.1° horizontal resolution (Parallel Ocean Program low and high resolutions [POP-LR and POP-HR], respectively)—are compared with the Argo data for 2005–2009. POP-LR has a shallow MLD bias common to many Fifth Coupled Models Intercomparison Project (CMIP5) models, while POP-HR has a mix of deep and shallow MLD biases. While both models were able to replicate the large-scale processes leading to formation of a high-salinity layer, the salinity feature in POP-HR is too strong and deep. Neither model was able to replicate the vertical mixing processes leading to penetration of the subsurface salt maximum.

54 ENVIRONMENTAL SCIENCES↗

DOE OSTI2023

Rapid discovery of high hardness multi-principal-element alloys using a generative adversarial network model

Multi-principal element alloys (MPEAs) continue to gain research prominence due to their promising high-temperature microstructural and mechanical properties. Recently, machine learning (ML) and materials informatics have been used extensively for screening MPEAs, however, most of these efforts were focused on constructing classification and regression models for predicting phase stability and mechanical properties of known compositions. These approaches may accelerate the screening process but optimizing new compositions with desirable properties within a practical time frame from an infinitely large design space of MPEA systems remains a grand challenge. To tackle this composition optimization challenge, a generative adversarial network coupled with a neural-network ML model was utilized to design MPEAs by filtering compositions that have high hardness. Even in a high-dimensional space with 18 elements as descriptors, the ML model was able to generate optimized compositions from which one composition was found to have 10% higher hardness (941 HV) than the maximum in the training data (857 HV). Density-functional theory was used to provide thermodynamic and electronic insights to higher hardness of the new MPEA found. The present work can optimize compositions from a wide design space of 18 elements (including W, Ta and Nb) that presents an opportunity to synthesize new compositions for applications ranging from corrosion-resistant alloys to nuclear materials. Here the findings suggest that generative ML can greatly accelerate materials discovery by identifying novel compositions, which can serve as a data-informed tool to guide experiments.

36 MATERIALS SCIENCE↗

NASA NTRS2004

The Nonlinear Coupling of Alfven and Lower Hybrid Waves in Space Plasma

Space plasmas support a wide variety of waves, and wave-particle interactions as well as wave-wave interactions which are of crucial importance to magnetospheric and ionospheric plasma behavior. The excitation of lower hybrid waves (LHWs) in particular is a widely discussed mechanism of interaction between plasma species in space and is one of the unresolved questions of magnetospheric multi-ion plasmas. It is demonstrated that large-amplitude Alfven waves may generate LHWs in the auroral zone and ring current region and in some cases (particularly in the inner magnetosphere) this serves as the Alfven wave saturation mechanism. We present several examples of observational data which illustrate that the proposed mechanism is a plausible candidate to explain certain classes of LHW generation events in the ionosphere and magnetosphere and demonstrate electron and ion energization involving these processes. We discuss the morphology dynamics and level of LHW activity generated by electromagnetic ion cyclotron (EMIC) waves during the May 2-7 1998 storm period on the global scale. The LHWs were calculated based on a newly developed self-consistent model (Khazanov et. al. 2002) that couples the system of two kinetic equations: one equation describes the ring current (RC) ion dynamic and another equation describes the evolution of EMIC waves. It is found that the LHWs are excited by helium ions due to their mass dependent drift in the electric field of EMIC waves. The level of LHW activity is calculated assuming that the induced scattering process is the main saturation mechanism for these waves. The calculated LHWs electric fields are consistent with the observational data.

Khazanov, George V.↗

DOE OSTI2019

Unsupervised machine learning based on non-negative tensor factorization for analyzing reactive-mixing

Analysis of reactive-diffusion simulations representing complex mixing processes requires a large number of independent model runs. For each high-fidelity model simulation, the model inputs are varied and the predicted mixing behavior is represented by temporal and spatial changes in species concentration. It is then required to discern how the model inputs (such as diffusivity, dispersion, anisotropy, and velocity field properties) impact the mixing process. This task is challenging and typically involves interpretation of large model outputs representing temporal and spatial changes of species concentration within the model domain. Yet, the task can be automated and substantially simplified by applying Machine Learning (ML) methods. Here, we present an application of an unsupervised ML method (called NTF k ) using Non-negative Tensor Factorization (NTF) coupled with a custom clustering procedure based on k-means to reveal the temporal and spatial features in product concentrations. An attractive and unique aspect of the proposed ML method is that it ensures the extracted features are non-negative, which is important to obtain a meaningful deconstruction of the mixing processes. The ML methodology is applied to a large set of high-resolution finite-element model simulations representing anisotropic reaction-diffusion processes in perturbed vortex-based velocity fields. The applied finite-element method ensures that spatial and temporal species concentration are always non-negative, even in the case of high anisotropic contrasts. The simulated reaction is a fast irreversible bimolecular reaction A+B→C , where species A and B react to form species C. The reactive-diffusion model input parameters that control mixing include properties of the velocity field (such as vortex structures), anisotropic dispersion, and molecular diffusion. We demonstrate the applicability of the ML feature extraction method to produce a meaningful deconstruction of model outputs to discriminate between different physical processes impacting the reactants, their mixing, and the spatial distribution of the product C. The introduced ML analysis allowed us to identify additive temporal and spatial features that characterize mixing behavior. The application of the proposed NTF k approach is not limited to reactive-mixing. NTFk can be readily applied to any observed or simulated datasets that can be represented as tensors (multi-dimensional arrays) and have separable latent signatures or features.

97 MATHEMATICS AND COMPUTING↗