Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “model structural error”

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 1,027 records · Page 57
All sourcesNASA NTRSDOE OSTI
DOE OSTI2022

GraphAlign: Graph-Enabled Machine Learning for Seismic Event Filtering

This report summarizes results from a 2 year effort to improve the current automated seismic event processing system by leveraging machine learning models that can operated over the inherent graph data structure of a seismic sensor network. Specifically, the GraphAlign project seeks to utilize prior information on which stations are more likely to detect signals originating from particular geographic regions to inform event filtering. To date, the GraphAlign team has developed a Graphical Neural Network (GNN) model to filter out false events generated by the Global Associator (GA) algorithm. The algorithm operates directly on waveform data that has been associated to an event by building a variable sized graph of station waveforms nodes with edge relations to an event location node. This builds off of previous work where random forest models were used to do the same task using hand crafted features. The GNN model performance was analyzed using an 8 week IMS/IDC dataset, and it was demonstrated that the GNN outperforms the random forest baseline. We provide additional error analysis of which events the GNN model performs well and poorly against concluded by future directions for improvements.

58 GEOSCIENCES↗
DOE OSTI2025

CALPHAD-based ICME design of single-step aging to enhance mechanical strength of WAAM Haynes 282

To match the strength of wire-arc additive manufactured Haynes 282 to its wrought counterpart via a single-step aging heat treatment, the CALPHAD (Calculation of Phase Diagrams) method is integrated with physics-based process-structure-property models and experimental validation. The integrated computational materials engineering (ICME) framework simulates the effects of aging on γ′ and M 23 C 6 precipitation and the resulting yield strength. To improve simulation reliability, the interfacial energies between γ/γ′ and γ/M 23 C 6 carbides were estimated by comparison with precipitation kinetic modeling and measured precipitate sizes. γ′ and M23C6 were found to precipitate simultaneously between 640 and 860 °C, producing microstructures similar to those produced by two-step aging. The optimal γ′ size for peak yield stress was calculated to be 20–23 nm. WAAM Haynes 282 aged at 780 °C for 50 h exceeded the mechanical performance of its wrought counterpart subjected to two-step aging, though desired properties can also be achieved at 800 °C for 16 h or less. The error in yield strength is less than 20 MPa, demonstrating good agreement between the modeling framework and experiments. Creep studies showed that WAAM Haynes 282 exceeded the calculated rupture time, reaching 481 h. This proposed methodology can accelerate the design of aging heat treatments for any γ′-strengthened nickel-base alloy, minimizing the resources required for trial-and-error experiments.

CALPHAD↗
DOE OSTI2022

Magnetic exchange interactions in binuclear and tetranuclear iron(III) complexes described by spin‐flip DFT and Heisenberg effective Hamiltonians

Abstract Low‐energy spectra of single‐molecule magnets (SMMs) are often described by Heisenberg Hamiltonians. Within this formalism, exchange interactions between magnetic centers determine the ground‐state multiplicity and energy separation between the ground and excited states. In this contribution, we extract exchange coupling constants ( J ) for a set of iron (III) binuclear and tetranuclear complexes from all‐electron calculations using non‐collinear spin‐flip time‐dependent density functional theory (NC‐SF‐TDDFT). For 12 binuclear complexes with J ‐values ranging from −6 to −132 cm −1 , our benchmark calculations using the short‐range hybrid ω PBEh functional and 6‐31G(d,p) basis set agree well with the experimentally derived values (mean absolute error of 4.7 cm −1 ). For the tetranuclear SMMs, the computed J constants are within 6 cm −1 from the experimentally derived values. We explore the range of applicability of the Heisenberg model by analyzing bonding patterns in these Fe(III) complexes using natural orbitals (NO), their occupations, and the number of effectively unpaired electrons. The results illustrate the efficiency of the spin‐flip protocol for computing the exchange couplings and the utility of the NO analysis in assessing the validity of effective spin Hamiltonians.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2025

A Block-Structured Adaptive Mesh Framework to Solve Radiation Transfer Equation in Irregular Embedded Geometries

Radiation transport arises in various scientific, industrial, and medical fields, and understanding its effect in applications is needed to make accurate predictions, safety assessments and performance optimizations. Solving the Radiation Transport Equation (RTE) is challenging due to its integro-differential nature, which involves both differential and integral terms. The differential term describes the change in radiation intensity due to absorption and emission, while the integral term accounts for scattering. The accurate modeling of radiation is further complicated in many applications due to the complex, irregular geometries. Various methods exist for solving the RTE, including the zonal, Monte Carlo, spherical harmonics, discrete ordinates, and finite volume methods. Traditional mesh-based approaches, which rely on structured or unstructured meshes, struggle with irregular geometries due to: a) the difficulty of conforming structured grids to irregular domains, b) challenges in enforcing boundary conditions correctly, and c) the additional computational cost of unstructured mesh methods. This work presents a second-order accurate method for solving the RTE in irregular geometries. The radiation intensity is discretized using the finite-volume method in both spatial and angular directions on regular Cartesian grid blocks. Leveraging the block-structured adaptive mesh refinement (AMR) framework provided by AMReX, our method refines the grid locally to reduce spatial discretization error, ensuring a converged numerical solution while minimizing computational costs elsewhere. A two-stage deferred correction approach is employed: First, a first-order discretization on grid blocks is solved using an algebraic multigrid method in HYPRE. Second, a correction term is applied explicitly to achieve second-order accuracy. The correction term is calculated by approximating the radiation flux on cell faces using a Total Variation Diminishing (TVD) scheme. This approach ensures quick convergence of the multigrid method while preserving higher-order accuracy of the numerical solution. Irregular geometries are resolved as embedded boundaries (EB), resulting in both cut cells and regular cells. In cut cells, we modify the fluxes using face fractions and incorporate additional contributions from EB boundary conditions. To ensure higher-order convergence near the EB interface, the correction term is modified by interpolating the radiation intensity to fictitious ghost points. The implementation takes advantage of modern supercomputers by leveraging AMReX’sMPI/X parallelization strategy where X can be MPI or a GPU accelerator including CUDA, HIP and DPC++. We validate our solver using classical test cases, both with and without EB, demonstrating accuracy and efficiency. Additionally, we analyze the impact of adaptive mesh refinement on solution accuracy and computational cost, highlighting the advantages of our approach for high-resolution radiation transport simulations.

computational fluid dynamics (CFD)↗
DOE OSTI2025

Robust Implicit Adaptive Low Rank Time-Stepping Methods for Matrix Differential Equations

In this work, we develop implicit rank-adaptive schemes for time-dependent matrix differential equations. The dynamic low rank approximation (DLRA) is a well-known technique to capture the dynamic low rank structure based on Dirac–Frenkel time-dependent variational principle. In recent years, it has attracted a lot of attention due to its wide applicability. Our schemes are inspired by the three-step procedure used in the rank adaptive version of the unconventional robust integrator (the so called BUG integrator) (Ceruti et al. in BIT Numer Math 62(4):1149–1174, 2022) for DLRA. First, a prediction (basis update) step is made computing the approximate column and row spaces at the next time level. Second, a Galerkin evolution step is invoked using an implicit solves for the small core matrix. Finally, a truncation is made according to a prescribed error threshold. Since the DLRA is evolving the differential equation projected on to the tangent space of the low rank manifold, the error estimate of the BUG integrator contains the tangent projection (modeling) error which cannot be easily controlled by mesh refinement. This can cause convergence issue for equations with cross terms. To address this issue, we propose a simple modification, consisting of merging the row and column spaces from the explicit step truncation method together with the BUG spaces in the prediction step. In addition, we propose an adaptive strategy where the BUG spaces are only computed if the residual for the solution obtained from the prediction space by explicit step truncation method, is too large. Here, we prove stability and estimate the local truncation error of the schemes under assumptions. We benchmark the schemes in several tests, such as anisotropic diffusion, solid body rotation and the combination of the two, to show robust convergence properties.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2022

$\mathrm{IH}$-$\mathrm{GAN}$: A conditional generative model for implicit surface-based inverse design of cellular structures

Variable-density cellular structures can overcome connectivity and manufacturability issues of topologically optimized structures, particularly those represented as discrete density maps. However, the optimization of such cellular structures is challenging due to the multiscale design problem. Past work addressing this problem generally either only optimizes the volume fraction of single-type unit cells but ignoring the effects of unit cell geometry on properties, or considers the geometry–property relation but builds this relation via heuristics. In contrast, we propose a simple yet more principled way to accurately model the property to geometry mapping using a conditional deep generative model, named Inverse Homogenization Generative Adversarial Network (IH-GAN). It learns the conditional distribution of unit cell geometries given properties and can realize the one-to-many mapping from properties to geometries. Here we further reduce the complexity of IH-GAN by using the implicit function parameterization to represent unit cell geometries. Results show that our method can 1) generate various unit cells that satisfy given material properties with high accuracy (R 2 -scores between target properties and properties of generated unit cells >98%) and 2) improve the optimized structural performance over the conventional variable-density single-type structure. In the minimum compliance example, our IH-GAN generated structure achieves a 79.7% reduction in concentrated stress and an extra 3.03% reduction in displacement. In the target deformation examples, our IH-GAN generated structure reduces the target matching error by 86.4% and 79.6% for two test cases, respectively. We also demonstrated that the connectivity issue for multi-type unit cells can be solved by transition layer blending.

42 ENGINEERING↗
DOE OSTI2024

A First-Principles Approach to Modeling Surface Site Stabilities on Multimetallic Catalysts

The study of multimetallic alloys and the multitude of possible surface compositions have sparked a tremendous interest in engineering low-cost materials with high activity and selectivity in heterogeneous catalysis. Multimetallic systems provide complementary functionalities and an unprecedented tunability when designing catalyst formulations. However, due to their immense structural and compositional complexity, the investigation and identification of an optimal catalyst is a tedious and time-consuming process, both experimentally and theoretically. Therefore, theoretical design principles are highly desirable to accelerate the screening of catalyst structures across the vast compositional space. In this paper, we introduce a simple and general model for predicting the site stability of multimetallic surfaces and nanoparticles, which is based on physical principles. The model requires only a small set of density functional theory (DFT) calculations of metal atom binding energies on monometallic and dilute alloy surface slabs to optimize the parameters in the simple model. The resulting model allows for the quantification of the stability of any particular atom site in any conceivable chemical environment across a wide range of morphologies, sizes, and arrangements by interpolating the derived parameters from a monometallic system to a completely diluted alloyed system. Herein, we demonstrate the robustness of the model across an extensive data set of transition metal alloy surfaces and 147-atoms cuboctahedral nanoparticles (NPs) composed of IrRhRu and PtPdRu. In conclusion, our approach yields mean absolute errors of ≈0.15 (IrRhRu), 0.20 (PtPdRu), 0.19 (IrRhRu NP), and 0.26 (PtPdRu NP) eV relative to site binding energies calculated using DFT.

36 MATERIALS SCIENCE↗
DOE OSTI2019

Thermodynamics and Kinetics of Phase Transformations in Energy Materials (Final Report)

The primary goal of this program is to develop computational tools and methodology for enabling a shift from empirical trial-and-error driven materials discovery towards data and information technology driven design of materials with targeted functional properties. Both thermodynamic and kinetic factors need to be taken into account when designing new materials. Thermodynamics sets the necessary-but-not-sufficient conditions for the suitability of a particular material in a given application, while the kinetics controls the rates of microscopic processes and hence the performance in energy applications (e.g., power in energy storage, conversion efficiencies in solar cells and thermoelectrics, chemical activity in catalysis). The complexity of these phenomena makes computational design difficult, not least due to the lack of suitable theoretical methods and computational tools. To address these needs, we develop systematic, quantitatively accurate first-principles methods to address two important problems: (i) predict and optimize the thermodynamics of materials over a large composition and structure space, and (ii) model kinetic processes such as mass and heat transport in crystalline solids under non-equilibrium conditions. These tasks are addressed by combining high-throughput computation with automatic construction of accurate lattice models using mathematically rigorous methods of information theory, such as compressive sensing and Bayesian inference.

36 MATERIALS SCIENCE↗
DOE OSTI2025

Equilibrium Core Model for Micro Pebble Bed Reactors Using OpenMC

Estimating the equilibrium state for pebble bed reactors (PBRs) presents complex challenges as it requires simultaneous consideration of changes in the pebbles’ movement as well as their fuel compositions. Whereas traditional approaches use multigroup diffusion codes for neutronics calculations of PBRs’ equilibrium state, the double-heterogeneity of PBRs complicates neutron cross-section generation. Continuous-energy Monte Carlo (MC) methods are better suited for detailed PBR analysis because of their natural handling of double-heterogeneity, but they demand substantially more computational resources. Here, this study introduces a novel method for efficiently estimating the equilibrium state in small and micro PBRs with reduced computational cost. The method is anticipated to accelerate the processes of core design and performing parametric studies for utilizing advanced fuel and structural materials. The HTR-10 reactor design was used for validating the method’s predictions and evaluating its computational efficiency. When compared to reference calculation values from the literature, criticality (k-effective) was predicted to be approximately within the margin of error of the MC transport calculation, average core power density (in megawatts per cubic meter) was predicted within 2.5% relative error, and maximum thermal flux (10 13 n/cm 2 .s −1 ) was predicted within 1.8% relative error. The calculated inventory of fission products and fuel composition in the equilibrium core were within 15% and 16.6%, respectively, when compared to reported values from the literature. The difference is attributed to variance in the considered values of the core temperature, which was found to significantly affect the depletion analyses.

Equilibrium core↗
DOE OSTI2015

Crystal structure representations for machine learning models of formation energies

Here we introduce and evaluate a set of feature vector representations of crystal structures for machine learning (ML) models of formation energies of solids. ML models of atomization energies of organic molecules have been successful using a Coulomb matrix representation of the molecule. We consider three ways to generalize such representations to periodic systems: (i) a matrix where each element is related to the Ewald sum of the electrostatic interaction between two different atoms in the unit cell repeated over the lattice; (ii) an extended Coulomb-like matrix that takes into account a number of neighboring unit cells; and (iii) an ansatz that mimics the periodicity and the basic features of the elements in the Ewald sum matrix using a sine function of the crystal coordinates of the atoms. The representations are compared for a Laplacian kernel with Manhattan norm, trained to reproduce formation energies using a dataset of 3938 crystal structures obtained from the Materials Project. For training sets consisting of 3000 crystals, the generalization error in predicting formation energies of new structures corresponds to (i) 0.49, (ii) 0.64, and (iii) urn:x-wiley:00207608:media:qua24917:qua24917-math-0001 for the respective representations.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2020

Quantifying Uncertainties In Neutron-Alpha Scattering With Chiral Nucleon-Nucleon And Three-Nucleon Forces

We report that modern ab initio theory combined with high-quality nucleon-nucleon (NN) and three-nucleon (3N) interactions from chiral effective field theory (EFT) can provide a predictive description of low-energy light-nuclei reactions relevant for astrophysics and fusion-energy applications. However, the high cost of computations has so far impeded a complete analysis of the uncertainty budget of such calculations. Starting from NN potentials up to fifth order (N 4 LO) combined with leading-order 3N forces, we study how the order-by-order convergence of the chiral expansion and confidence intervals for the 3N contact and contact-plus-one-pion-exchange low-energy constants (c E and c D ) contribute to the overall uncertainty budget of many-body calculations of neutron- 4 He (n-α) elastic scattering. We compute structure and reaction observables for three-, four-, and five-nucleon systems within the ab initio frameworks of the no-core shell model and no-core shell model with continuum. Using a small set of design runs, we construct a Gaussian process model (GPM) that acts as a statistical emulator for the theory. With this, we gain insight into how uncertainties in the 3N low-energy constants propagate throughout the calculation and determine the Bayesian posterior distribution of these parameters with Markov-Chain Monte Carlo. We find rapidly converging n-α phase shifts with respect to the chiral order. With the adopted leading-order 3N force, calculations based on the NN interaction at N 4 LO of Entem, Machleidt, and Nosyk are unable to reproduce the experimental phase shifts in the 3/2 - channel within the estimated chiral truncation errors. Closer agreement with empirical data is found when using an older parametrization of the NN interaction at order N 3 LO, and the position and width of the P-wave resonances can be used to reduce the uncertainty of the 3N low-energy constants. The present results point to a lack of spin-orbit strength when the newer parametrization of the chiral NN force up to fifth order is combined with the leading-order 3N force. The inclusion of higher-order 3N-force terms may be required to recover the missing strength. GPMs can act as fast and accurate emulators of ab initio many-body calculations of low-energy scattering and reactions of light nuclei, opening the way to a robust quantification of theoretical uncertainties grounded in the description of the underlying chiral Hamiltonian.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
DOE OSTI2019

Cluster Cosmology with the Velocity Distribution Function of the HeCS-SZ Sample

We apply the velocity distribution function (VDF) to a sample of Sunyaev–Zel'dovich (SZ)-selected clusters, and we report preliminary cosmological constraints in the ${\sigma }_{8}$-${{\rm{\Omega }}}_{m}$ cosmological parameter space. The VDF is a forward-modeled test statistic that can be used to constrain cosmological models directly from galaxy cluster dynamical observations. The method was introduced in Ntampaka et al. and employs line-of-sight velocity measurements to directly constrain cosmological parameters; it is less sensitive to measurement error than a standard halo mass function approach. The method is applied to the Hectospec Survey of Sunyaev–Zeldovich-Selected Clusters sample, which is a spectroscopic follow-up of a Planck-selected sample of 83 galaxy clusters. Credible regions are calculated by comparing the VDF of the observed cluster sample to that of mock observations, yielding ${{ \mathcal S }}_{8}$ $\equiv \,{\sigma }_{8}{\left({{\rm{\Omega }}}_{m}/0.3\right)}^{0.25}=0.751\pm 0.037$. These constraints are in tension with the Planck Cosmic Microwave Background TT fiducial value, which lies outside of our 95% credible region, but are in agreement with some recent analyses of large-scale structure that observe fewer massive clusters than are predicted by the Planck fiducial cosmological parameters.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2019

The Aemulus Project. II. Emulating the Halo Mass Function

Existing models for the dependence of the halo mass function on cosmological parameters will become a limiting source of systematic uncertainty for cluster cosmology in the near future. We present a halo mass function emulator and demonstrate improved accuracy relative to state-of-the-art analytic models. In this work, mass is defined using an overdensity criteria of 200 relative to the mean background density. Our emulator is constructed from the AEMULUS simulations, a suite of 40 N-body simulations with snapshots from z = 3 to z = 0. These simulations cover the flat wCDM parameter space allowed by recent cosmic microwave background, baryon acoustic oscillation and SNe Ia results, varying the parameters w, Omega(m), Omega(b), sigma(8), N-eff, n(s), and H-0. We validate our emulator using five realizations of seven different cosmologies, for a total of 35 test simulations. These test simulations were not used in constructing the emulator, and were run with fully independent initial conditions. We use our test simulations to characterize the modeling uncertainty of the emulator, and introduce a novel way of marginalizing over the associated systematic uncertainty. We confirm nonuniversality in our halo mass function emulator as a function of both cosmological parameters and redshift. Our emulator achieves better than 1% precision over much of the relevant parameter space, and we demonstrate that the systematic uncertainty in our emulator will remain a negligible source of error for cluster abundance studies through at least the LSST Year 1 data set.

large-scale structure of universe↗
DOE OSTI2022

Reduced order modeling for flow and transport problems with Barlow Twins self-supervised learning

Abstract We propose a unified data-driven reduced order model (ROM) that bridges the performance gap between linear and nonlinear manifold approaches. Deep learning ROM (DL-ROM) using deep-convolutional autoencoders (DC–AE) has been shown to capture nonlinear solution manifolds but fails to perform adequately when linear subspace approaches such as proper orthogonal decomposition (POD) would be optimal. Besides, most DL-ROM models rely on convolutional layers, which might limit its application to only a structured mesh. The proposed framework in this study relies on the combination of an autoencoder (AE) and Barlow Twins (BT) self-supervised learning, where BT maximizes the information content of the embedding with the latent space through a joint embedding architecture. Through a series of benchmark problems of natural convection in porous media, BT–AE performs better than the previous DL-ROM framework by providing comparable results to POD-based approaches for problems where the solution lies within a linear subspace as well as DL-ROM autoencoder-based techniques where the solution lies on a nonlinear manifold; consequently, bridges the gap between linear and nonlinear reduced manifolds. We illustrate that a proficient construction of the latent space is key to achieving these results, enabling us to map these latent spaces using regression models. The proposed framework achieves a relative error of 2% on average and 12% in the worst-case scenario (i.e., the training data is small, but the parameter space is large.). We also show that our framework provides a speed-up of $$7 \times 10^{6}$$ 7 × 10 6 times, in the best case, and $$7 \times 10^{3}$$ 7 × 10 3 times on average compared to a finite element solver. Furthermore, this BT–AE framework can operate on unstructured meshes, which provides flexibility in its application to standard numerical solvers, on-site measurements, experimental data, or a combination of these sources.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2025

Beyond real: alternative unitary cluster Jastrow models for molecular electronic structure calculations on near-term quantum computers

Near-term quantum devices require wavefunction ansätze that are expressive while also of shallow circuit depth in order to both accurately and efficiently simulate molecular electronic structure. While the unitary coupled cluster ansatz (e.g., UCCSD) has become a standard, the high gate count associated with the implementation of this limits its feasibility on noisy intermediate-scale quantum (NISQ) hardware. k -Fold unitary cluster Jastrow (uCJ) ansätze mitigate this challenge by providing O( kN 2 ) circuit scaling and favorable linear depth circuit implementation. Previous work has focused on the real orbitalrotation (Re-uCJ) variant of uCJ, which allows an exact (Trotter-free) implementation. Here we extend and generalize the k -fold uCJ framework by introducing two new variants, Im-uCJ and g-uCJ, which incorporate imaginary and fully complex orbital rotation operators, respectively. Similar to Re-uCJ, both of the new variants achieve quadratic gate-count scaling. Our results focus on the simplest k = 1 model, and show that the uCJ models frequently maintain energy errors within chemical accuracy (∼1 kcal mol −1 ). Both g-uCJ and Im-uCJ are more expressive in terms of capturing electron correlation and are also more accurate than the earlier Re-uCJ ansatz. We further show that Im-uCJ and g-uCJ circuits can also be implemented exactly, without any Trotter decomposition. Numerical tests using k = 1 on H 2 , H 3 + , Be 2 , C 2 H 4 , C 2 H 6 and C 6 H 6 in various basis sets confirm the practical feasibility of these shallow Jastrow-based ansätze for applications on near-term quantum hardware.

Tkachenko, Nikolay V. [University of California, B↗
DOE OSTI2014

A new polarizable force field for alkali and halide ions

We developed transferable potentials for alkali and halide ions which are consistent with our recent model of water [P. T. Kiss and A. Baranyai, J. Chem. Phys. 138, 204507 (2013)]. Following the approach used for the water potential, we applied Gaussian charge distributions, exponential repulsion, and r{sup −6} attraction. One of the two charges of the ions is fixed to the center of the particle, while the other is connected to this charge by a harmonic spring to express polarization. Polarizability is taken from quantum chemical calculations. The repulsion between different species is expressed by the combining rule of Kong [J. Chem. Phys. 59, 2464 (1972)]. Our primary target was the hydration free energy of ions which is correct within the error of calculations. We calculated water-ion clusters up to 6 water molecules, and, as a crosscheck, we determined the density and internal energy of alkali-halide crystals at ambient conditions with acceptable accuracy. The structure of hydrated ions was also discussed.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2023

Di-CNN: Domain-Knowledge-Informed Convolutional Neural Network for Manufacturing Quality Prediction

In manufacturing, convolutional neural networks (CNNs) are widely used on image sensor data for data-driven process monitoring and quality prediction. However, as purely data-driven models, CNNs do not integrate physical measures or practical considerations into the model structure or training procedure. Consequently, CNNs’ prediction accuracy can be limited, and model outputs may be hard to interpret practically. This study aims to leverage manufacturing domain knowledge to improve the accuracy and interpretability of CNNs in quality prediction. A novel CNN model, named Di-CNN, was developed that learns from both design-stage information (such as working condition and operational mode) and real-time sensor data, and adaptively weighs these data sources during model training. It exploits domain knowledge to guide model training, thus improving prediction accuracy and model interpretability. A case study on resistance spot welding, a popular lightweight metal-joining process for automotive manufacturing, compared the performance of (1) a Di-CNN with adaptive weights (the proposed model), (2) a Di-CNN without adaptive weights, and (3) a conventional CNN. The quality prediction results were measured with the mean squared error (MSE) over sixfold cross-validation. Model (1) achieved a mean MSE of 6.8866 and a median MSE of 6.1916, Model (2) achieved 13.6171 and 13.1343, and Model (3) achieved 27.2935 and 25.6117, demonstrating the superior performance of the proposed model.

47 OTHER INSTRUMENTATION↗