Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Model Validation”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

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

At least 973 records · Page 54

Dynamic Modeling for the Separation of Rare Earth Elements Using Solvent Extraction: Predicting Separation Performance Using Laboratory Equilibrium Data

Industrial rare earth element (REE) separation facilities utilize acidic cation exchange ligands such as 2-ethylhexylphosphonic acid mono-2-ethylhexyl ester (PC88A) for solvent extraction processes. REE separations are costly and difficult due to their chemical similarities and subsequent low separation factors. Several empirical correlations are available in the literature to predict steady state extraction equilibria for various solvent systems. However, complete solvent extraction flowsheet design for REE separations requires complex scrubbing and stripping circuits to separate and produce individual pure species. Furthermore, the sensitive process chemistry is susceptible to process upsets and may manifest as irreversible transients that complicate startup and continuous operation of a solvent extraction plant. Thus, dynamic modeling of REE separations will aid in process design, optimization, and management of process fluctuations. A time-dependent MATLAB/SIMULINK modeling framework has been developed for REE solvent extraction processes using laboratory equilibrium data. The model was used to develop a flowsheet that produced high purity neodymium from a 25 wt% praseodymium and 75 wt% neodymium feed. Laboratory mixer-settlers were used to verify and validate model performance. Results indicated that the model reasonably predicts the dynamic behavior of a counter-current REE separation process, and accurately predicts the steady state REE concentration profiles across the cascade. The modeling framework can be applied to any REE separation or solvent system provided adequate laboratory equilibrium data are available.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Cluster dynamics models of irradiation damage accumulation in ferritic iron. I. Trap mediated interstitial cluster diffusion

The microstructure that develops under low temperature irradiation in ferritic alloys is dominated by a high density of small (2–5 nm) defects. These defects have been widely observed to move via occasional discrete hops during in situ thin film irradiation experiments. Cluster dynamics models are used to describe the formation of these defects as an aggregation process of smaller clusters created as primary damage. Multiple assumptions regarding the mobility of these damage features are tested in the models, both with and without explicit consideration of such irradiation induced hops. Comparison with experimental data regarding the density of these defects demonstrates the importance of including such motions in a valid model. In particular, discrete hops inform the limited dependence of defect density on irradiation temperature observed in experiments, which the model was otherwise incapable of producing.

36 MATERIALS SCIENCE↗

Understanding and predicting profile structure and parametric scaling of intrinsic rotation

This study reports on a recent advance in developing physical understanding and a first-principles-based model for predicting intrinsic rotation profiles in magnetic fusion experiments. It is shown for the first time that turbulent fluctuation-driven residual stress (a non-diffusive component of momentum flux) along with diffusive momentum flux can account for both the shape and magnitude of the observed intrinsic toroidal rotation profile. Both the turbulence intensity gradient and zonal flow E×B shear are identified as major contributors to the generation of the k ∥ -asymmetry needed for the residual stress generation. The model predictions of core rotation based on global gyrokinetic simulations agree well with the experimental measurements of main ion toroidal rotation for a set of DIII-D ECH discharges. The validated model is further used to investigate the characteristic dependence of residual stress and intrinsic rotation profile structure on the multi-dimensional parametric space covering the turbulence type, q-profile structure, and up-down asymmetry in magnetic geometry with the goal of developing the physics understanding needed for rotation profile control and optimization. It is shown that in the flat-q profile regime, intrinsic rotations driven by ITG and TEM turbulence are in the opposite direction (i.e., intrinsic rotation reverses). The predictive model also produces reversed intrinsic rotation for plasmas with weak and normal shear q-profiles.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Consistent evaluation of GOSAT, SCIAMACHY, carbontracker, and MACC through comparisons to TCCON

Consistent validation of satellite CO 2 estimates is a prerequisite for using multiple satellite CO 2 measurements for joint flux inversion, and for establishing an accurate long-term atmospheric CO 2 data record. We focus on validating model and satellite observation attributes that impact flux estimates and CO 2 assimilation, including accurate error estimates, correlated and random errors, overall biases, biases by season and latitude, the impact of coincidence criteria, validation of seasonal cycle phase and amplitude, yearly growth, and daily variability. We evaluate dry air mole fraction (X CO 2 ) for GOSAT (ACOS b3.5) and SCIAMACHY (BESD v2.00.08) as well as the CarbonTracker (CT2013b) simulated CO 2 mole fraction fields and the MACC CO 2 inversion system (v13.1) and compare these to TCCON observations (GGG2014). We find standard deviations of 0.9 ppm, 0.9, 1.7, and 2.1 ppm versus TCCON for CT2013b, MACC, GOSAT, and SCIAMACHY, respectively, with the single target errors 1.9 and 0.9 times the predicted errors for GOSAT and SCIAMACHY, respectively. When satellite data are averaged and interpreted according to error 2 = a 2 + b 2 / n (where n are the number of observations averaged, a are the systematic (correlated) errors, and b are the random (uncorrelated) errors), we find that the correlated error term a = 0.6 ppm and the uncorrelated error term b = 1.7 ppm for GOSAT and a = 1.0 ppm, b = 1.4 ppm for SCIAMACHY regional averages. Biases at individual stations have year-to-year variability of ~ 0.3 ppm, with biases larger than the TCCON predicted bias uncertainty of 0.4 ppm at many stations. Using fitting software, we find that GOSAT underpredicts the seasonal cycle amplitude in the Northern Hemisphere (NH) between 46–53° N. In the Southern Hemisphere (SH), CT2013b underestimates the seasonal cycle amplitude. Biases are calculated for 3-month intervals and indicate the months that contribute to the observed amplitude differences. The seasonal cycle phase indicates whether a dataset or model lags another dataset in time. We calculate this at a subset of stations where there is adequate satellite data, and find that the GOSAT retrieved phase improves substantially over the prior and the SCIAMACHY retrieved phase improves substantially for 2 of 7 sites. The models reproduce the measured seasonal cycle phase well except for at Lauder125 (CT2013b), Darwin (MACC), and Izana (+ 10 days, CT2013b), as for Bremen and Four Corners, which are highly influenced by local effects. We compare the variability within one day between TCCON and models in JJA; there is correlation between 0.2 and 0.8 in the NH, with models showing 10–100 % the variability of TCCON at different stations (except Bremen and Four Corners which have no variability compared to TCCON) and CT2013b showing more variability than MACC. This paper highlights findings that provide inputs to estimate flux errors in model assimilations, and places where models and satellites need further investigation, e.g. the SH for models and 45–67° N for GOSAT

54 ENVIRONMENTAL SCIENCES↗

Learning continuous models for continuous physics

Abstract Dynamical systems that evolve continuously over time are ubiquitous throughout science and engineering. Machine learning (ML) provides data-driven approaches to model and predict the dynamics of such systems. A core issue with this approach is that ML models are typically trained on discrete data, using ML methodologies that are not aware of underlying continuity properties. This results in models that often do not capture any underlying continuous dynamics—either of the system of interest, or indeed of any related system. To address this challenge, we develop a convergence test based on numerical analysis theory. Our test verifies whether a model has learned a function that accurately approximates an underlying continuous dynamics. Models that fail this test fail to capture relevant dynamics, rendering them of limited utility for many scientific prediction tasks; while models that pass this test enable both better interpolation and better extrapolation in multiple ways. Our results illustrate how principled numerical analysis methods can be coupled with existing ML training/testing methodologies to validate models for science and engineering applications.

97 MATHEMATICS AND COMPUTING↗

From deep TLS validation to ensembles of atomic models built from elemental motions. II. Analysis of TLS refinement results by explicit interpretation

TLS modelling was developed by Schomaker and Trueblood to describe atomic displacement parameters through concerted (rigid-body) harmonic motions of an atomic group [Schomaker & Trueblood (1968), Acta Cryst. B24, 63–76]. The results of a TLS refinement are T, L and S matrices that provide individual anisotropic atomic displacement parameters (ADPs) for all atoms belonging to the group. These ADPs can be calculated analytically using a formula that relates the elements of the TLS matrices to atomic parameters. Alternatively, ADPs can be obtained numerically from the parameters of concerted atomic motions corresponding to the TLS matrices. Both procedures are expected to produce the same ADP values and therefore can be used to assess the results of TLS refinement. Here, the implementation of this approach in PHENIX is described and several illustrations, including the use of all models from the PDB that have been subjected to TLS refinement, are provided.

59 BASIC BIOLOGICAL SCIENCES↗

A simplified model of TiH1.65/KClO4 pyrotechnic ignition.

A simplified model was developed and is presented in this report for simulating thermal transport coupled with chemical reactions that lead to the pyrotechnic ignition of TiH1.65/KClO4 powder. The model takes into account Joule heating via a bridgewire, thermal contact resistance at the wire/powder interface, convective heat loss to the surroundings, and heat released from the TiH1.65- and KClO4-decomposition and TiO2-oxidation reactions. Chemical kinetic sub-models were put forth to describe the chemical reaction rate(s) and quantify the resultant heat release. The simplified model predicts pyrotechnic ignition when heat from the pyrotechnic reactions is accounted for. Effects of six key parameters on ignition were examined. It was found that the two reaction-rate parameters and the thermal contact resistance significantly affect the dynamic ignition process whereas the convective heat transfer coefficient essentially has no effect on the ignition time. Effects of the initial/ambient temperature and electrical current load through the wire are as expected. Ignition time increases as the initial/ambient temperature is lowered or the wire current load is reduced. Lastly, critical needs such as experiments to determine reaction-rate and other model-input parameters and to measure temperature profiles, time to ignition and burn-rate data for model validation as well as efforts in incorporating reaction-rate dependency on pressure are pointed out.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Uncertainties in model-independent extractions of amplitudes from complete experiments

A new generation of over-complete experiments is underway, with the goal of performing a high precision extraction of pseudoscalar meson photo-production amplitudes. Such experimentally determined amplitudes can be used both as a test to validate models and as a starting point for an analytic continuation in the complex plane to search for poles. Of crucial importance for both is the level of uncertainty in the extracted multipoles. We have probed these uncertainties by analyses of pseudo-data for KLambda photoproduction, first for the set of 8 observables that have been published for the K{sup +} Lambda channel and then for pseudo-data on a complete set of 16 observables with the uncertainties expected from analyses of ongoing CLAS experiments. In fitting multipoles, we have used a combined Monte Carlo sampling of the amplitude space, with gradient minimization, and have found a shallow X{sup 2} valley pitted with a large number of local minima. This results in bands of solutions that are experimentally indistinguishable. All ongoing experiments will measure observables with limited statistics. We have found a dependence on the particular random choice of values of Gaussian distributed pseudo-data, due to the presence of multiple local minima. This results in actual uncertainties for reconstructed multipoles that are often considerable larger than those returned by gradient minimization routines such as Minuit which find a single local minimum. As intuitively expected, this additional level of uncertainty decreases as larger numbers of observables are included.

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗

A Benchmark Study on Casting Residual Stress

Stringent regulatory requirements, such as Tier IV norms, have pushed the cast iron for automotive applications to its limit. The castings need to be designed with closer tolerances by incorporating hitherto unknowns, such as residual stresses arising due to thermal gradients, phase and microstructural changes during solidification phenomenon. Residual stresses were earlier neglected in the casting designs by incorporating large factors of safety. Experimental measurement of residual stress in a casting through neutron or X-ray diffraction, sectioning or hole drilling, magnetic, electric or photoelastic measurements is very difficult and time consuming exercise. A detailed multi-physics model, incorporating thermo-mechanical and phase transformation phenomenon, provides an attractive alternative to assess the residual stresses generated during casting. However, before relying on the simulation methodology, it is important to rigorously validate the prediction capability by comparing it to experimental measurements. In the present work, a benchmark study was undertaken for casting residual stress measurements through neutron diffraction, which was subsequently used to validate the accuracy of simulation prediction. The stress lattice specimen geometry was designed such that subsequent castings would generate adequate residual stresses during solidification and cooling, without any cracks. The residual stresses in the cast specimen were measured using neutron diffraction. Considering the difficulty in accessing the neutron diffraction facility, these measurements can be considered as benchmark for casting simulation validations. Simulations were performed using the identical specimen geometry and casting conditions for predictions of residual stresses. The simulation predictions were found to agree well with the experimentally measured residual stresses. The experimentally validated model can be subsequently used to predict residual stresses in different cast components. This enables incorporation of the residual stresses at the design phase along with external loads for accurate predictions of fatigue and fracture performance of the cast components.

36 MATERIALS SCIENCE↗

A novel machine learning based identification of potential adopter of rooftop solar photovoltaics

With the proliferation of rooftop solar photovoltaic installations, there is a need to proactively predict consumer potential for solar photovoltaic adoption, for improved electric utility planning and operation. Traditional analytical modeling approaches are limited to a few survey features and a larger part of the survey would remain untouched by the decision model. This article presents a novel, data-driven modeling approach that strategically prunes a large set of consumer profile features using a machine learning framework to train a model for predicting potential solar adoption. The approach utilizes the Gradient Boosting Decision Tree model through a Light Gradient Boosting framework that improves significantly over the poor prediction accuracy of the existing approaches. Model training using focal-loss based supervision is used to overcome the difficulty in identifying the potential adopters that is inherent in conventional data-driven models. In addition, to overcome possible data sparsity in a limited survey sample, a Generative Adversarial Network is presented to create synthetic user samples and its effectiveness on model performance is assessed. A Bayesian optimization approach is used to systematically arrive at the hyperparameters of the proposed model. Validation of the presented approach on a survey data collected by the National Rural Electric Cooperative Association in Virginia in 2018 demonstrates the excellent predictive capability of the machine learning based approach to modeling solar adoption reliably.

14 SOLAR ENERGY↗

Effects of various parameters of different porous transport layers in proton exchange membrane water electrolysis

Porous transport layers (PTLs) play an important role in proton exchange membrane water electrolysis (PEMWE) cells. The PTL facilitates water and gas transport, as well as thermal and electrical conduction, and is required to sustain good contact with adjacent components. It is expected that using PTLs with variations in material properties such as structure, composition, thickness and wettability results in performance changes of the PEMWE. Here, a general mathematical PEMWE model is developed that separates and analyzes the contributions of ohmic, activation, diffusion and Nernst potentials. For model validation, three inherently different anode PTL structures (carbon paper, sintered titanium particles, and titanium felt) are operated over a range of conditions. Additionally, the effects of PTL wettability were used to verify the model using Polytetrafluoroethylene (PTFE) treated Toray papers with PTFE loading ranging from 0% to 20%. The modeling results of both PTFE treated and untreated materials show good agreement with the experimental data. Mass transport or diffusion loss is the primary reason for performance differences between PTFE treated and untreated PTLs. Sintered titanium PTLs with thicknesses above 1 mm suffer from up to 33% increased ohmic losses without indicating any obvious changes in activation and diffusion losses when compared to untreated PTLs. The losses of the cell increase when using PTFE treated Toray paper. Individual contributions are quantified and assigned to increased ohmic, activation, and diffusion losses. In conclusion, the proposed model offers insights into the overpotential contributions of a PEMWE. It is a useful tool for predicting performance of various PTL materials and can be applied for PTL development and optimization efforts.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Fuel properties to enable lifted-flame combustion

The Fuel Properties to Enable Lifted-Flame Combustion project responded directly to solicitation DE-FOA-0000239 AOI 1A, Fuels and Lubricants for Advanced Combustion Regimes. This subtopic was intended to encompass clean and highly-efficient, liquid-fueled combustion engines to achieve extremely low engine-out nitrogen oxides (NOx) and particulate matter (PM) as a target and similar efficiency as state-of-the-art direct injection diesel engines. The intent of this project was to identify how fuel properties can be used to achieve controllable Leaner Lifted Flame Combustion (LLFC) with low NOx and PM emissions. Specifically, this project was expected to identify and test key fuel properties to enable LLFC and their compatibility with current fuel systems and to enhance combustion models to capture the effect of fuel properties on advanced combustion. Successful demonstration of LLFC may reduce the need for after treatment devices, thereby reducing costs and improving thermal efficiency. The project team consisted of key technical personnel from Ford Motor Company (FMC), the University of Wisconsin-Madison (UW), Sandia National Laboratories (SNL) and Lawrence Livermore National Laboratories (LLNL). Each partner had key roles in achieving project objectives. FMC investigated fuel properties relating to LLFC and sooting tendency. Together, FMC and UW developed and integrated 3D combustion models to capture fuel property combustion effects. FMC used these modeling results to develop a combustion system and define fuel properties to support a single-cylinder demonstration of fuel-enabled LLFC. UW investigated modeling the flame characteristics and emissions behavior of different fuels, including those with different cetane number and oxygen content. SNL led spray combustion experiments to quantify the effect of key fuel properties on combustion characteristics critical for LLFC, as well as single cylinder optical engine experiments to improve fundamental understanding of flame lift-off, generate model validation data, and demonstrate LLFC concurrent with FMC efforts. Additionally, LLNL was added to the project during the second year to develop a detailed kinetic mechanism for a key oxygenate to support CFD modeling. Successful completion of this project allowed the team to enhance fundamental understanding of LLFC, improve the state of current combustion models and increase understanding of desired fuel properties. This knowledge also improves our knowledge of how cost effective and environmentally friendly renewable fuels can assist in helping meet future emission and greenhouse gas regulations.

Kurtz, Eric↗

FY-2018 FES Joint Research Target

Fusion alphas and injected energetic neutral particle beams provide an important source of heating and current drive in advanced tokamak operating scenarios and burning plasma regimes. Alfvén eigenmode instabilities can cause the redistribution or loss of fast ions and driven currents, as well as potentially decreasing fusion performance and leading to localized losses. Measured fast ion fluxes in DIII-D and NSTX-U plasmas with different levels of Alfvén eigenmode activity will be used to determine the threshold for significant fast ion transport, assess mechanisms and models for such transport, and quantify the impact on beam power deposition and current drive. Measurements will be compared with theoretical predictions, including quantitative fluctuation data and fast ion density, in order to validate models and improve understanding of underlying mechanisms. Model predictions will guide the development of attractive operating regimes.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Development and validation of a model for tritium accumulation by a freshwater bivalve using the IAEA EMRAS scenarios

A six-compartment metabolic model for tritium accumulation by bivalves was developed and validated using two observed data sets supplied in an international IAEA program for validation of environmental models, EMRAS (Environmental Modeling for Radiation Safety, 2003-2007). The data observed were presented in scenarios for model prediction of temporal change of HTO and OBT concentrations in Barnes mussels (Elliptio complanata). In the Uptake Scenario, mussels were transplanted from a site with background tritium concentrations into a lake, which has historically received tritium inputs over time from up-gradient waste management areas. Another data set was presented in the Depuration Scenario for model prediction of the temporal decrease in HTO and OBT concentrations in the mussels following transplantation from the lake into another lake with significantly lower tritium levels. The model simulation was able to reproduce the observation that the amount of hydrogen taken from sediment was very small compared with that taken from lake water. (authors)

54 ENVIRONMENTAL SCIENCES↗

Benchmarking Initiatives

Benchmarking has been recognized as an efficient tool for model validation as well as scientific collaboration. Several workshop series have been established to foster the benchmarking idea: setting up test cases with increasing complexity for method development and code comparison. In addition to representing model complexity, one of the key efforts is to develop codes that are suitable for modern HPC platforms such as PetaFlop supercomputers. It has been realized that those challenges are beyond single team capabilities. The Subsurface Environmental Simulation Benchmarking (SSBench) working group was initiated to develop, publish, and make publicly available well-described benchmark problems that could be used to demonstrate simulator conformance with norms established by the subsurface science and engineering community. SSBench supports the development of environmentally-relevant benchmark problem sets that provide rigorous tests for process model conceptualization, numerical implementation, process coupling, and accuracy. The initial SSBench focus on multicomponent reactive transport modeling reflects the expanding range of mechanisms and increasing complexity needed to address issues relevant to climate change, nuclear waste management, contaminant remediation, and pollution prevention.

benchmarking↗

Data-driven emulation of modal aerosol microphysics via neural operator-based modeling

The complexity and the small characteristic scales of aerosol microphysical processes pose a big challenge for accurate and efficient Earth system simulations at regional and global scales. In this work, we construct and evaluate a surrogate model: the aerosol deep operator network (ADON), a physics-inspired dual-net architecture for emulating the aerosol microphysics parameterization suite in the version 2 of the Energy Earth System Model (E3SMv2). The current version of the surrogate model is trained on a dataset comprising 9.8 million samples obtained from a global E3SMv2 simulation with the horizontal resolution of about one degree under cloud-free conditions. Incorporating domain spatial and temporal coordinates, as well as principle components extracted from training data, the dual-net surrogate model effectively captures the intricate representations of aerosol and the relationship with atmospheric state variables, achieving an R-squared score over $$95.7\%$$ for all the lognormal aerosol modes in the extrapolated regime. The validated model provides feature importance of input variables and their impact on the predictive capacity of the surrogate model in relation to the E3SM. The computational cost of online inference time deployed on CPUs and GPUs with lower precisions highlights ADON’s efficiency and potential in robust predictive modeling for large-scale Earth system computations.

Bai, Zhe↗

Continuum Model to Define the Chemistry and Mass Transfer in a Bicarbonate Electrolyzer

Bicarbonate electrolyzers are devices designed to convert CO 2 captured from point sources or the atmosphere into chemicals and fuels without needing to first isolate pure CO 2 gas. In this work, we report here an experimentally validated model that quantifies the reaction chemistry and mass transfer processes within the catalyst layer and cation exchange membrane layer of a bicarbonate electrolyzer. Our results demonstrate that two distinct chemical microenvironments are key to forming CO at high rates: an acidic membrane layer that promotes in situ CO 2 formation and a basic catalyst layer that suppresses the hydrogen evolution reaction. We show that the rate of CO product formation can be increased by modulating the catalyst and membrane layer properties to increase the rate of in situ CO 2 generation and transport to the cathode. These insights serve to inform the design of bicarbonate and BPM-based CO 2 electrolyzers while demonstrating the value of modeling for resolving rate-determining processes in electrochemical systems.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Modeling and Simulation Tool to Improve Urban Search Planning (NA22 Final Report)

The venture project Tool to Improve Urban Search Planning (LL14-V-Urban Search Planning ToolPD3RS) supported two major types of effort: 1) development of a software prototype search planning tool, with the goal of transitioning the tool to operational users for evaluation and possible adoption; 2) enabling research to better understand the radiation characteristics of an urban environment, with implications for deployment of radiation detection systems in that environment. The first effort has resulted in prototype software, Optimization Planning Tool for Urban Search (OPTUS), that has been distributed to a limited set of potential users, including users within the NNSA emergency response community. The second effort supported various research topics including high fidelity radiation transport modeling in urban environments, measurement campaigns for model validation, development of vehicle-based high purity germanium (HPGe) detector systems and supporting data analysis, and computational capabilities for archiving and managing large data sets. While some of these related efforts have informed the models and software design of the OPTUS tool, in large part they pursued more basic research, some of which has evolved into separate DNN R&D projects. Project performers included Oak Ridge National Laboratory (ORNL), Lawrence Berkeley National Laboratory (LBNL) and the Remote Sensing Laboratory (RSL), in addition to Lawrence Livermore Laboratory (LLNL), the project lead. The project also supported university research, including several internships at the laboratories.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗