Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “level set methods”

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

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

At least 703 records · Page 39

DeepUQ: Assessing the Aleatoric Uncertainties from two Deep Learning Methods

Assessing the quality of aleatoric uncertainty estimates from uncertainty quantification (UQ) deep learning methods is important in scientific contexts, where uncertainty is physically meaningful and important to characterize and interpret exactly. We systematically compare aleatoric uncertainty measured by two UQ techniques, Deep Ensembles (DE) and Deep Evidential Regression (DER). Our method focuses on both zero-dimensional (0D) and two-dimensional (2D) data, to explore how the UQ methods function for different data dimensionalities. We investigate uncertainty injected on the input and output variables and include a method to propagate uncertainty in the case of input uncertainty so that we can compare the predicted aleatoric uncertainty to the known values. We experiment with three levels of noise. The aleatoric uncertainty predicted across all models and experiments scales with the injected noise level. However, the predicted uncertainty is miscalibrated to $\rm{std}(\sigma_{\rm al})$ with the true uncertainty for half of the DE experiments and almost all of the DER experiments. The predicted uncertainty is the least accurate for both UQ methods for the 2D input uncertainty experiment and the high-noise level. While these results do not apply to more complex data, they highlight that further research on post-facto calibration for these methods would be beneficial, particularly for high-noise and high-dimensional settings.

Nevin, Rebecca↗

Oscillator strengths and transition probabilities for allowed and forbidden transitions in Fe XIX

An extensive set of oscillator strengths, line strengths, and radiative decay rates for the allowed and forbidden transitions in Fe XIX is presented. They correspond to 1626 fine structure levels of total angular momenta 0{<=}J{<=}8 of even and odd parities with 2{<=}n{<=}10, 0{<=}l{<=}9, 0{<=}L{<=}10, and (2S+1)=1, 3, 5. In contrast, the compiled table of the National Institute for Standards and Technology (NIST) lists only 63 observed levels. A total of 289,291 electric dipole allowed transitions are presented. They were obtained in the close coupling approximation using the relativistic Breit-Pauli R-matrix method. The wavefunction expansion included 15 levels of the configurations 2s{sup 2}2p{sup 3}, 2s2p{sup 4}, and 2p{sup 5} of the Fe XX core. The calculated fine structure levels are assigned with spectroscopic identifications using quantum defect analysis. Comparison with the observed energies shows very good agreement, the largest difference being less than 4%. The transitions also compare well with the compiled data by NIST and recent calculations. The forbidden transitions of the electric quadrupole and octupole, and magnetic dipole and quadrupole, type are presented for the 379 levels of the configurations 2s{sup 2}2p{sup 4}, 2s2p{sup 5}, 2p{sup 6}, 2s{sup 2}2p{sup 3}3s, 2s{sup 2}2p{sup 3}3p, 2s{sup 2}2p{sup 3}3d, 2s{sup 2}2p{sup 3}4s, 2s{sup 2}2p{sup 3}4p, 2s{sup 2}2p{sup 3}4d, 2s{sup 2}2p{sup 3}4f, 2s2p{sup 4}3s, 2s2p{sup 4}3p, 2s2p{sup 4}3d, 2s2p{sup 4}4s, 2s2p{sup 4}4p, and 2s{sup 2}2p{sup 2}3s{sup 2} of Fe XIX. They correspond to a total of 66,619 transitions. These results have been obtained from relativistic Breit-Pauli atomic structure calculations using the program SUPERSTRUCTURE. The forbidden transition probabilities show very good agreement with those compiled by NIST. - Highlights: {yields} Presents the most complete (n up to 10) set of transitions for Fe XIX. {yields} Considers both allowed and forbidden transitions. {yields} Large number of new energy levels are found. {yields} Benchmarks existing accurate energies and transitions. {yields} Completeness and accuracy are crucial to solve its abundance in the Sun.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Modeling underwater noise propagation from marine hydrokinetic power devices through a time-domain, velocity-pressure solution

Marine hydrokinetic (MHK) devices generate electricity from the motion of tidal and ocean currents, as well as ocean waves, to provide an additional source of renewable energy available to the United States. These devices are a source of anthropogenic noise in the marine ecosystem and must meet regulatory guidelines that mandate a maximum amount of noise that may be generated. In the absence of measured levels from in situ deployments, a model for predicting the propagation of sound from an array of MHK sources in a real environment is essential. A set of coupled, linearized velocity-pressure equations in the time-domain are derived and presented in this paper, which are an alternative solution to the Helmholtz and wave equation methods traditionally employed. Discretizing these equations on a three-dimensional (3D), finite-difference grid ultimately permits a finite number of complex sources and spatially varying sound speeds, bathymetry, and bed composition. The solution to this system of equations has been parallelized in an acoustic-wave propagation package developed at Sandia National Labs, called Paracousti. This work presents the broadband sound pressure levels from a single source in two-dimensional (2D) ideal and Pekeris wave-guides and in a 3D domain with a sloping boundary. Furthermore, the paper concludes with demonstration of Paracousti for an array of MHK sources in a simple wave-guide.

13 HYDRO ENERGY↗

Risk-Informed Safety Margin Characterization Case Study: Selection of Electrical Equipment to Be Subjected to Environmental Qualification

The Risk-Informed Safety Margin Characterization (RISMC) pathway of the DOE’s Light Water Reactor Sustainability (LWRS) program focuses on advancing the state of the art in safety analysis and risk assessment to support decision-making on nuclear power plant operation well beyond the originally designed lifetime of the plants (i.e., beyond 60 years). Among the issues being addressed in RISMC is the significance of SSC aging and how confident we are about our understanding of its impact on the margin between the loads SSCs are expected to see during normal operation and accident conditions, and the SSC capacities (their ability to resist those loads) as the SSCs age. In this paper, a summary is provided of a case study that examines SSC aging from an environmental qualification (EQ) perspective. The case study illustrates how the state of knowledge regarding SSC margin can be characterized given the overall integrated plant design, and was developed to demonstrate a method for deciding on which cables to focus, which cables are not so important from an environmental qualification margin standpoint, and what plant design features or operating characteristics determine the role that environmental qualification plays in establishing a safety case on which decisions regarding margin can be made. The selection of cables for which demonstration of margin with respect to aging and environmental challenges uses a technique known as Prevention Analysis. Prevention Analysis is a Boolean method for optimal selection of SSCs (that is, those combinations of SSCs both necessary and sufficient to meet a predetermined selection criterion) in a manner that allows demonstration that plant-level safety can be demonstrated by the collection of selected SSCs alone. Choosing the set of SSCs that is necessary and sufficient to satisfy the safety objectives, and demonstrating that the safety objectives can be met effectively, determines where resources are best allocated to assure SSC performance margin. The paper describes the resulting component types that were selected by Prevention Analysis and identifies the accident sequence characteristics that cause these component types to be important from an EQ and aging perspective (and, hence, worthwhile evaluating the extent of safety margin). In addition, component types not selected as needing significant margin from an EQ and aging perspective are discussed and an engineering rationale is developed justifying the lack of need to apply resources to demonstrating margin for these component types. This rationale is in terms of design features of the plant and operating characteristics that make these component types less important from an EQ and aging perspective. While the case study focuses on EQ and aging of equipment and cables located inside the containment of this PWR, the prevention analysis method is demonstrated to be an effective technique for identification of minimal collections of components that would be effective in managing safety for a variety of issues associated with aging and long-term operation of the fleet of plants.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

Collisional and Radiative Effects in Transient sub-Doppler Hole Burning: Double Resonance Measurements in CN

We report transient hole-burning and saturation recovery measurements in the CN radical with MHz frequency resolution and 20 ns time resolution. Narrow velocity groups of individual hyperfine levels of selected rotational states in CN (X{sup 2} {Sigma}{sup +}) are depleted and excited (A{sup 2}{pi}{sub i}) with a saturation laser and probed by a counterpropagating, frequency modulated probe beam. Recent work in our lab has used this method to measure and characterize the hyperfine splittings for a set of rotational, fine structure, and parity components of CN (A{sup 2}{pi}{sub i}, v=1). Extending this work, we report time and frequency dependence of the saturation signals following abrupt switching of the CW saturation beam on and off with an electro-optic amplitude modulator. Recovery of the unsaturated absorption following the turnoff of the saturation beam follows pressure-dependent kinetics, driven by collisions with the undissociated NCCN precursor with a rate coefficient of 2 x 10{sup -9} cm{sup 3} s{sup -1} molec{sup -1}. Similar recovery kinetics are observed for two-level saturation resonances, where the signal observed is a combination of X- and A-state kinetics, as well as for three-level crossover resonances, which can be chosen to probe selectively the holefilling in the X state or the decay of velocity-selected A state radicals. The observed recovery rates are 8-10 times faster than the estimated rotationally inelastic contribution. The observed recovery rates are likely dominated by velocity-changing collisions in both X and A states, occurring with similar rates, despite the large difference in the properties of these electronic states. Transient signal risetimes following the turning on of the saturation pulse are consistent with the expected Rabi frequency. At lower pressures ({approx}50 mTorr) and higher beam power ({approx}200 mW), we can observe multiple Rabi cycles before collisions disrupt the coherent excitation and the transient signal reaches a steady state.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Computerized detection of noncalcified plaques in coronary CT angiography: Evaluation of topological soft gradient prescreening method and luminal analysis

Purpose: The buildup of noncalcified plaques (NCPs) that are vulnerable to rupture in coronary arteries is a risk for myocardial infarction. Interpretation of coronary CT angiography (cCTA) to search for NCP is a challenging task for radiologists due to the low CT number of NCP, the large number of coronary arteries, and multiple phase CT acquisition. The authors conducted a preliminary study to develop machine learning method for automated detection of NCPs in cCTA. Methods: With IRB approval, a data set of 83 ECG-gated contrast enhanced cCTA scans with 120 NCPs was collected retrospectively from patient files. A multiscale coronary artery response and rolling balloon region growing (MSCAR-RBG) method was applied to each cCTA volume to extract the coronary arterial trees. Each extracted vessel was reformatted to a straightened volume composed of cCTA slices perpendicular to the vessel centerline. A topological soft-gradient (TSG) detection method was developed to prescreen for NCP candidates by analyzing the 2D topological features of the radial gradient field surface along the vessel wall. The NCP candidates were then characterized by a luminal analysis that used 3D geometric features to quantify the shape information and gray-level features to evaluate the density of the NCP candidates. With machine learning techniques, useful features were identified and combined into an NCP score to differentiate true NCPs from false positives (FPs). To evaluate the effectiveness of the image analysis methods, the authors performed tenfold cross-validation with the available data set. Receiver operating characteristic (ROC) analysis was used to assess the classification performance of individual features and the NCP score. The overall detection performance was estimated by free response ROC (FROC) analysis. Results: With our TSG prescreening method, a prescreening sensitivity of 92.5% (111/120) was achieved with a total of 1181 FPs (14.2 FPs/scan). On average, six features were selected during the tenfold cross-validation training. The average area under the ROC curve (AUC) value for training was 0.87 ± 0.01 and the AUC value for validation was 0.85 ± 0.01. Using the NCP score, FROC analysis of the validation set showed that the FP rates were reduced to 3.16, 1.90, and 1.39 FPs/scan at sensitivities of 90%, 80%, and 70%, respectively. Conclusions: The topological soft-gradient prescreening method in combination with the luminal analysis for FP reduction was effective for detection of NCPs in cCTA, including NCPs causing positive or negative vessel remodeling. The accuracy of vessel segmentation, tracking, and centerline identification has a strong impact on NCP detection. Studies are underway to further improve these techniques and reduce the FPs of the CADe system.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Tunable and angle-insensitive plasmon resonances in graphene ribbon arrays with multispectral diffraction response

Plasmon resonances in graphene ribbon arrays are investigated numerically by means of the Finite Element Method. Numerical analysis shows that a series of multipolar resonances take place when graphene ribbon arrays are illuminated by a TM polarized electromagnetic wave. Moreover, these resonances are angle-independent, and can be tuned greatly by the width and the doping level of the graphene ribbons. Specifically, we demonstrate that for graphene arrays with several sets of graphene ribbons, which have different widths or doping levels, each of these multipolar resonances will be split into several ones. In addition, as plasmon resonances can confine electromagnetic field at the ribbon edges, graphene ribbons with different widths or doping levels offer intriguing application for electrically tunable spectral imaging.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Multi-fidelity modeling to predict the rheological properties of fiber suspensions

Unveiling the rheological properties of fiber suspensions is of paramount interest to many industrial applications like biofuel production. The 3D numerical simulations of the suspension of fibers are often computationally expensive and time-consuming. Machine learning methods such as neural networks can simplify the prediction of rheological behavior; however, they require a relatively large training data set. Multi -fidelity models, which combine high-fidelity data from numerical simulations and less expensive lower fidelity data from resources such as simplified physical equations, can lead to optimized predictions. Here, we focus on a neural network with two levels of fidelity, i.e., high and low fidelity networks. To produce high-fidelity data, we perform direct numerical simulations to model the fibers as one-dimensional inextensible slender bodies that obey the Euler- Bernoulli beam equation. The Navier-Stokes equations govern the suspended fluid, and an immersed boundary method is used to couple the fluid and solid motion. The low-fidelity data is produced by using constitutive equations. Noticeable improvements have been observed in the accuracy of predicting the rheological behavior when a multi-fidelity network is used compared to the single-fidelity network.

Boodaghidizaji, Miad↗

The quantitative potential for breast tomosynthesis imaging

Purpose: Due to its limited angular scan range, breast tomosynthesis has lower resolution in the depth direction, which may limit its accuracy in quantifying tissue density. This study assesses the quantitative potential of breast tomosynthesis using relatively simple reconstruction and image processing algorithms. This quantitation could allow improved characterization of lesions as well as image processing to present tomosynthesis images with the familiar appearance of mammography by preserving more low-frequency information. Methods: All studies were based on a Siemens prototype MAMMOMAT Novation TOMO breast tomo system with a 45 deg. total angular span. This investigation was performed using both simulations and empirical measurements. Monte Carlo simulations were conducted using the breast tomosynthesis geometry and tissue-equivalent, uniform, voxelized phantoms with cuboid lesions of varying density embedded within. Empirical studies were then performed using tissue-equivalent plastic phantoms which were imaged on the actual prototype system. The material surrounding the lesions was set to either fat-equivalent or glandular-equivalent plastic. From the simulation experiments, the effects of scatter, lesion depth, and background material density were studied. The empirical experiments studied the effects of lesion depth, background material density, x-ray tube energy, and exposure level. Additionally, the proposed analysis methods were independently evaluated using a commercially available QA breast phantom (CIRS Model 11A). All image reconstruction was performed with a filtered backprojection algorithm. Reconstructed voxel values within each slice were corrected to reduce background nonuniformities. Results: The resulting lesion voxel values varied linearly with known glandular fraction (correlation coefficient R{sup 2}>0.90) under all simulated and empirical conditions, including for the independent tests with the QA phantom. Analysis of variance performed on the fit line parameters revealed statistically significant differences between the two different background materials and between 28 kVp and the remaining energies (26, 30, and 32 kVp) for the dense experimental phantom. However, no significant differences arose between different energies for the fatty phantom, nor for any of the many other combinations of parameters. Conclusions: These strong linear relationships suggest that breast tomosynthesis image voxel values, after being corrected by our outlined methods, are highly positively correlated with true tissue density. This consistent linearity implies that breast tomosynthesis imaging indeed has potential to be quantitative.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

A modified homotopy method for solving the principal-agent bilevel programming problem

In this paper, the principal-agent bilevel programming problem with integral operator is considered, in which the upper-level object is the agent that maximizes its expected utility with respect to an agreed compensation contract. The constraints are the principal’s participation and the agent’s incentive compatibility. The latter is a lower-level optimization problem with respect to its private action. To solve an equivalent single-level nonconvex programming problem with integral operator, a modified homotopy method for solving the Karush–Kuhn–Tucker system is proposed. This method requires only an interior point and, not necessarily, a feasible initial approximation for the constraint shifting set. Global convergence is proven under some mild conditions. Numerical experiments were performed by our homotopy method as well as by fmincon in Matlab, LOQO and MINOS. The experiments showed that: designing a piecewise linear contract is much better than designing a piecewise constant contract and only needs to solve a much lower-dimensional optimization problem and hence needs much less computation time; the optimal value of the principal-agent model with designing piecewise linear contract tends to a limitation, while the discrete segments gradually increase; and finally, the proposed modified homotopy method is feasible and effective.

97 MATHEMATICS AND COMPUTING↗

Calculations for energies, transition rates, and lifetimes in Al-like Kr XXIV

Using the second-order many-body perturbation theory (MBPT) method, a complete and accurate data set of excitation energies, lifetimes, wavelengths, and electric dipole (E1), magnetic dipole (M1), electric quadrupole (E2), and magnetic quadrupole (M2) line strengths, transition rates, and oscillator strengths for the lowest 880 levels arising from the 3l{sup 3} (0≤l≤2), 3l{sup 2}4l{sup ′} (0≤l≤2, 0≤l{sup ′}≤3), 3s{sup 2}5l (0≤l≤4), 3p{sup 2}5l (0≤l≤1), and 3s3p5l (0≤l≤4) configurations in Al-like Kr XXIV is provided. Comparisons are made with available experimental and theoretical results. Our calculated energies are expected to be accurate enough to facilitate identifications of observed lines involving the n=4,5 levels. The complete data set is also useful for modeling and diagnosing fusion plasma.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Parallel Computation of Coupled Cluster Hyperpolarizabilities

Static hyperpolarizabilities of five benchmark systems (hydrogen fluoride, water, acetonitrile, chloroform and para-nitroaniline) are calculated with large basis sets using coupled-cluster response theory and compared to five common DFT methods. These results reveal which methods and basis sets are appropriate for nonlinear optical studies for different types of molecules and provide a means for estimating errors from the quantum chemical approximation when including vibrational contributions or solvent effects at the QM/MM level. The largest calculation reported, which was for 72 electrons in 812 functions at C2v symmetry, took only a few hours on 256 nodes demonstrating that even larger calculations are quite feasible using modern supercomputers.

08 HYDROGEN↗

Could the description on polynuclear superhalogens by DFT be comparable with high-level ab initio results? A comparison between DFT and CCSD(T)

A systematic density functional theory study including 17 exchange-correlation functionals was performed on different types of superhalogens with high level coupled-cluster single double including perturbative triple excitations (CCSD(T)) results as the reference. The superhalogens selected here cover the ranges from mononuclear to polynuclear structures and from structures with halogen-atom ligands to those with non-halogen ligands, e.g., [MgX{sub 3}]{sup −}, [Mg{sub 2}X{sub 5}]{sup −}, and [Mg{sub 3}X{sub 7}]{sup −} (X = F, Cl, CN). It is clearly indicated that three double-hybrid functionals B2T-PLYP, B2GP-PLYP, B2K-PLYP as well as the range-separated hybrid functional ωB97X are capable of providing results which approach the accuracy at the CCSD(T) level. The basis set effect is usually moderate and, in most cases, it is enough to utilize the basis set of triple-ξ quality, e.g., Def2-TZVP. In addition, the results of the HF and MP2 method are also acceptable here, especially for polynuclear superhalogens where CCSD(T) is probably unpractical.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Extraction of spin-dependent parton densities and their uncertainties

We discuss techniques and results for the extraction of the nucleon's spin-dependent parton distributions and their uncertainties from data for polarized deep-inelastic lepton-nucleon and proton-proton scattering by means of a global QCD analysis. Computational methods are described that significantly increase the speed of the required calculations to a level that allows one to perform the full analysis consistently at next-to-leading order accuracy. We examine how the various data sets help to constrain different aspects of the quark, antiquark, and gluon helicity distributions. Uncertainty estimates are performed using both the Lagrange multiplier and the Hessian approaches. We use the extracted parton distribution functions and their estimated uncertainties to predict spin asymmetries for high-transverse momentum pion and jet production in polarized proton-proton collisions at 500 GeV center-of-mass system energy at the Relativistic Heavy Ion Collider at Brookhaven National Laboratory, as well as for W boson production.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Initial Considerations for Large-Scale Carbon Removal in the United States

The Biden Administration has established a goal of removing 1 Gt carbon dioxide equivalents per year (CO 2e /yr) from the atmosphere and of achieving net zero greenhouse gas emissions (GHG) by 2050. Our team of leading academic and DOE national laboratory experts is conducting the first economy-wide technical evaluation of the options for achieving this carbon dioxide removal (CDR) goal. The net zero goal includes targets of 100% clean electricity by 2035 with 40% GHG emissions reductions by 2030, which helps set primary scenario boundaries for our analysis of the CDR supply curve. Our analysis will evaluate feasibility, performance, and costs on a county level for the entire USA (including AK and HI where possible), considering all removal methods that are currently well-enough developed for us to estimate the likely costs in 2050. We anticipate that more than 1 Gt CO 2e of removal will be available to the Nation. We will identify how much of each CO 2 removal approach is available in specific regions of the Nation and provide cumulative costs and volumes (a supply curve) by region for 2050. We expect to complete this detailed analysis by late 2023.

42 ENGINEERING↗

Heap/stack guard pages using a wakeup unit

A method and system for providing a memory access check on a processor including the steps of detecting accesses to a memory device including level-1 cache using a wakeup unit. The method includes invalidating level-1 cache ranges corresponding to a guard page, and configuring a plurality of wakeup address compare (WAC) registers to allow access to selected WAC registers. The method selects one of the plurality of WAC registers, and sets up a WAC register related to the guard page. The method configures the wakeup unit to interrupt on access of the selected WAC register. The method detects access of the memory device using the wakeup unit when a guard page is violated. The method generates an interrupt to the core using the wakeup unit, and determines the source of the interrupt. The method detects the activated WAC registers assigned to the violated guard page, and initiates a response.

97 MATHEMATICS AND COMPUTING↗

Python and HPC for High Energy Physics Data Analyses

High level abstractions in Python that can utilize computing hardware well seem to be an attractive option for writing data reduction and analysis tasks. In this paper, we explore the features available in Python which are useful and efficient for end user analysis in High Energy Physics (HEP). A typical vertical slice of an HEP data analysis is somewhat fragmented: the state of the reduction/analysis process must be saved at certain stages to allow for selective reprocessing of only parts of a generally time-consuming workflow. Also, algorithms tend to to be modular because of the heterogeneous nature of most detectors and the need to analyze different parts of the detector separately before combining the information. This fragmentation causes difficulties for interactive data analysis, and as data sets increase in size and complexity (O10 TiB for a “small” neutrino experiment to the O10 PiB currently held by the CMS experiment at the LHC), data analysis methods traditional to the field must evolve to make optimum use of emerging HPC technologies and platforms. Mainstream big data tools, while suggesting a direction in terms of what can be done if an entire data set can be available across a system and analysed with high-level programming abstractions, are not designed with either scientific computing generally, or modern HPC platform features in particular, such as data caching levels, in mind. Our example HPC use case is a search for a new elementary particle which might explain the phenomenon known as “Dark Matter”. Here, using data from the CMS detector, we will use HDF5 as our input data format, and MPI with Python to implement our use case.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Fast Evaluation of Aircraft Icing Severity Using Machine Learning Based on XGBoost

Aircraft icing represents a serious hazard in aviation which has caused a number of fatal accidents over the years. In addition, it can lead to substantial increase in drag and weight, thus reducing the aerodynamics performance of the airplane. The process of ice accretion on a solid surface is a complex interaction of aerodynamic and environmental variables. The complex relationship makes machine learning-based methods an attractive alternative to traditional numerical simulation-based approaches. In this study, we introduce a purely data-driven approach to find the complex pattern between different flight conditions and aircraft icing severity prediction. The supervised learning algorithm Extreme Gradient Boosting (XGBoost) is applied to establish the prediction framework which makes prediction based on any set of observations. The input flight conditions for the proposed prediction framework are liquid water content, droplet diameter and exposure time. The proposed approach is demonstrated in three cases: maximum ice thickness prediction, icing area prediction and icing severity level evaluation. Performance comparison studies and error analysis are also conducted to verify the effectiveness and performance of the proposed method. Results show that the proposed method has reasonable capability in evaluating aircraft icing severity.

Li, Sibo (ORCID:000000021705844X)↗