This study presents the development and characterization of a novel white reflective filament suitable for additive manufacturing of finely segmented plastic scintillators using 3D printing. The filament is based on polycarbonate (PC) and polymethyl methacrylate (PMMA) polymers loaded with titanium dioxide (TiO$_2$) and polytetrafluoroethylene (PTFE) to enhance reflectivity. A range of filament compositions and thicknesses was evaluated through optical reflection and transmittance measurements. Reflective layers were made by using the Fused Deposition Modeling (FDM) technique. A 3D-segmented plastic scintillator prototype was made with fused injection modeling (FIM) and tested with cosmic rays to assess the light yield and the optical crosstalk. The results demonstrate the feasibility of producing compact and modular 3D-printed scintillator detectors with a performance analogous to standard plastic scintillator detectors, with lower light crosstalk, thus higher light yield, compared to past works, owing to the improved optical properties of the reflector material.
The semiconductor industry faces a computational crisis in extreme ultraviolet (EUV) lithography optimization, where traditional methods consume billions of CPU hours while failing to achieve sub-nanometer precision. We present a physics-constrained adaptive learning framework that automatically calibrates electromagnetic approximations through learnable parameters $\boldsymbolθ = \{θ_d, θ_a, θ_b, θ_p, θ_c\}$ while simultaneously minimizing Edge Placement Error (EPE) between simulated aerial images and target photomasks. The framework integrates differentiable modules for Fresnel diffraction, material absorption, optical point spread function blur, phase-shift effects, and contrast modulation with direct geometric pattern matching objectives, enabling cross-geometry generalization with minimal training data. Through physics-constrained learning on 15 representative patterns spanning current production to future research nodes, we demonstrate consistent sub-nanometer EPE performance (0.664-2.536 nm range) using only 50 training samples per pattern. Adaptive physics learning achieves an average improvement of 69.9\% over CNN baselines without physics constraints, with a significant inference speedup over rigorous electromagnetic solvers after training completion. This approach requires 90\% fewer training samples through cross-geometry generalization compared to pattern-specific CNN training approaches. This work establishes physics-constrained adaptive learning as a foundational methodology for real-time semiconductor manufacturing optimization, addressing the critical gap between academic physics-informed neural networks and industrial deployment requirements through joint physics calibration and manufacturing precision objectives.
This paper presents a novel online transfer learning approach in state-based potential games (TL-SbPGs) for distributed self-optimization in manufacturing systems. The approach targets practical industrial scenarios where knowledge sharing among similar players enhances learning in large-scale and decentralized environments. TL-SbPGs enable players to reuse learned policies from others, which improves learning outcomes and accelerates convergence. To accomplish this goal, we develop transfer learning concepts and similarity criteria for players, which offer two distinct settings: (a) predefined similarities between players and (b) dynamically inferred similarities between players during training. The applicability of the SbPG framework to transfer learning is formally established. Furthermore, we present a method to optimize the timing and weighting of knowledge transfer. Experimental results from a laboratory-scale testbed show that TL-SbPGs improve production efficiency and reduce power consumption compared to vanilla SbPGs.
Mikhail Khrenov, Moon Tan, Lauren Fitzwater
et al.
Metal additive manufacturing (AM) opens the possibility for spatial control of as-fabricated microstructure and properties. However, since the solid state diffusional transformations that drive microstructure outcomes are governed by nonlinear ODEs in terms of temperature, which is itself governed by PDEs over the entire part domain, solving for the system inputs needed to achieve desired microstructure distributions has proven difficult. In this work, we present a trajectory optimization approach for spatial control of microstructure in metal AM, which we demonstrate by controlling the hardness of a low-alloy steel in electron beam powder bed fusion (EB-PBF). To this end, we present models for thermal and microstructural dynamics. Next, we use experimental data to identify the parameters of the microstructure transformation dynamics. We then pose spatial microstructure control as a finite-horizon optimal control problem. The optimal power field trajectory is computed using an augmented Lagrangian differential dynamic programming (AL-DDP) method with GPU acceleration. The resulting time-varying power fields are then realized on an EB-PBF machine through an approximation scheme. Measurements of the resultant hardness shows that the optimized power field trajectory is able to closely produce the desired hardness distribution.
Bioresorbable Mg-based alloys with low density, low elastic modulus, and excellent biocompatibility are outstanding candidates for temporary orthopedic implants. Coincidentally, metal additive manufacturing (AM) is disrupting the biomedical sector by providing fast access to patient-customized implants. Due to the high cooling rates associated with fusion-based AM techniques, they are often described as rapid solidification processes. However, conclusive observations or rapid solidification in metal AM -- attested by drastic microstructural changes induced by solute trapping, kinetic undercooling, or morphological transitions of the solid-liquid interface -- are scarce. Here we study the formation of banded microstructures during laser powder-bed fusion (LPBF) of a biomedical-grade Magnesium-rare earth alloy, combining advanced characterization and state-of-the-art thermal and phase-field modeling. Our experiments unambiguously identify microstructures as the result of an oscillatory banding instability known from other rapid solidification processes. Our simulations confirm that LPBF-relevant solidification conditions strongly promote the development of banded microstructures in a Mg-Nd alloy. Simulations also allow us to peer into the sub-micrometer nanosecond-scale details of the solid-liquid interface evolution giving rise to the distinctive banded patterns. Since rapidly solidified Mg alloys may exhibit significantly different mechanical and corrosion response compared to their cast counterparts, the ability to predict the emergence of rapid solidification microstructures (and to correlate them with local solidification conditions) may open new pathways for the design of bioresorbable orthopedic implants, not only fitted geometrically to each patient, but also optimized with locally-tuned mechanical and corrosion properties.
An innovative method is developed for accurate determination of thermodynamic properties as a function of temperature by revisiting the density functional theory (DFT) based quasiharmonic approach (QHA). The present methodology individually evaluates the contributions from static total energy, phonon, and thermal electron to free energy for increased efficiency and accuracy. The Akaike information criterion with a correction (AICc) is used to select models and model parameters for fitting each contribution as a function of volume. Using the additively manufactured Inconel alloy 625 (IN625) as an example, predicted temperature-dependent linear coefficient of thermal expansion (CTE) agrees well with dilatometer measurements and values in the literature. Sensitivity and uncertainty are also analyzed for the predicted IN625 CTE due to different structural configurations used by DFT, and hence different equilibrium properties determined.
D. Tourret, J. Klemm-Toole, A. Eres Castellanos
et al.
Understanding rapid solidification behavior at velocities relevant to additive manufacturing (AM) is critical to controlling microstructure selection. Although in-situ visualization of solidification dynamics is now possible, systematic studies under AM conditions with microstructural outcomes compared to solidification theory remain lacking. Here we measure solid-liquid interface velocities of Ni-Mo-Al alloy single crystals under AM conditions with synchrotron X-ray imaging, characterize the microstructures, and show discrepancies with classical theories regarding the onset velocity for absolute stability of a planar solid-liquid interface. Experimental observations reveal cellular/dendritic microstructures can persist at velocities larger than the expected absolute stability limit, where banded structure formation should theoretically appear. We show that theory and experimental observations can be reconciled by properly accounting for the effect of solute trapping and kinetic undercooling on the velocity-dependent solidus and liquidus temperatures of the alloy. Further theoretical developments and accurate assessments of key thermophysical parameters - like liquid diffusivities, solid-liquid interface excess free energies, and kinetic coefficients - remain needed to quantitatively investigate such discrepancies and pave the way for the prediction and control of microstructure selection under rapid solidification conditions.
Richard Nordsieck, André Schweizer, Michael Heider
et al.
Procedural knowledge describes how to accomplish tasks and mitigate problems. Such knowledge is commonly held by domain experts, e.g. operators in manufacturing who adjust parameters to achieve quality targets. To the best of our knowledge, no real-world datasets containing process data and corresponding procedural knowledge are publicly available, possibly due to corporate apprehensions regarding the loss of knowledge advances. Therefore, we provide a framework to generate synthetic datasets that can be adapted to different domains. The design choices are inspired by two real-world datasets of procedural knowledge we have access to. Apart from containing representations of procedural knowledge in Resource Description Framework (RDF)-compliant knowledge graphs, the framework simulates parametrisation processes and provides consistent process data. We compare established embedding methods on the resulting knowledge graphs, detailing which out-of-the-box methods have the potential to represent procedural knowledge. This provides a baseline which can be used to increase the comparability of future work. Furthermore, we validate the overall characteristics of a synthesised dataset by comparing the results to those achievable on a real-world dataset. The framework and evaluation code, as well as the dataset used in the evaluation, are available open source.
Deploying deep learning-based applications in specialized domains like the aircraft production industry typically suffers from the training data availability problem. Only a few datasets represent non-everyday objects, situations, and tasks. Recent advantages in research around Vision Foundation Models (VFM) opened a new area of tasks and models with high generalization capabilities in non-semantic and semantic predictions. As recently demonstrated by the Segment Anything Project, exploiting VFM's zero-shot capabilities is a promising direction in tackling the boundaries spanned by data, context, and sensor variety. Although, investigating its application within specific domains is subject to ongoing research. This paper contributes here by surveying applications of the SAM in aircraft production-specific use cases. We include manufacturing, intralogistics, as well as maintenance, repair, and overhaul processes, also representing a variety of other neighboring industrial domains. Besides presenting the various use cases, we further discuss the injection of domain knowledge.
Transfer learning (TL) based additive manufacturing (AM) modeling is an emerging field to reuse the data from historical products and mitigate the data insufficiency in modeling new products. Although some trials have been conducted recently, the inherent challenges of applying TL in AM modeling are seldom discussed, e.g., which source domain to use, how much target data is needed, and whether to apply data preprocessing techniques. This paper aims to answer those questions through a case study defined based on an open-source dataset about metal AM products. In the case study, five TL methods are integrated with decision tree regression (DTR) and artificial neural network (ANN) to construct six TL-based models, whose performances are then compared with the baseline DTR and ANN in a proposed validation framework. The comparisons are used to quantify the performance of applied TL methods and are discussed from the perspective of similarity, training data size, and data preprocessing. Finally, the source AM domain with larger qualitative similarity and a certain range of target-to-source training data size ratio are recommended. Besides, the data preprocessing should be performed carefully to balance the modeling performance and the performance improvement due to TL.
Francis Ogoke, Kyle Johnson, Michael Glinsky
et al.
Laser Powder Bed Fusion has become a widely adopted method for metal Additive Manufacturing (AM) due to its ability to mass produce complex parts with increased local control. However, AM produced parts can be subject to undesirable porosity, negatively influencing the properties of printed components. Thus, controlling porosity is integral for creating effective parts. A precise understanding of the porosity distribution is crucial for accurately simulating potential fatigue and failure zones. Previous research on generating synthetic porous microstructures have succeeded in generating parts with high density, isotropic porosity distributions but are often inapplicable to cases with sparser, boundary-dependent pore distributions. Our work bridges this gap by providing a method that considers these constraints by deconstructing the generation problem into its constitutive parts. A framework is introduced that combines Generative Adversarial Networks with Mallat Scattering Transform-based autocorrelation methods to construct novel realizations of the individual pore geometries and surface roughness, then stochastically reconstruct them to form realizations of a porous printed part. The generated parts are compared to the existing experimental porosity distributions based on statistical and dimensional metrics, such as nearest neighbor distances, pore volumes, pore anisotropies and scattering transform based auto-correlations.
Harald Garcke, Kei Fong Lam, Robert Nürnberg
et al.
A phase field approach for structural topology optimization with application to additive manufacturing is analyzed. The main novelty is the penalization of overhangs (regions of the design that require underlying support structures during construction) with anisotropic energy functionals. Convex and non-convex examples are provided, with the latter showcasing oscillatory behavior along the object boundary termed the dripping effect in the literature. We provide a rigorous mathematical analysis for the structural topology optimization problem with convex and non-continuously-differentiable anisotropies, deriving the first order necessary optimality condition using subdifferential calculus. Via formally matched asymptotic expansions we connect our approach with previous works in the literature based on a sharp interface shape optimization description. Finally, we present several numerical results to demonstrate the advantages of our proposed approach in penalizing overhang developments.
Andreas Rumsch, Christoph Imboden, Alberto Calatroni
et al.
More and more household appliances connect to the Internet and exchange data freely. This is the foundation for true smart buildings. However, there is still no uniform communication technology available, which can connect all appliances from all vendors. Protocols differ between manufacturers making interoperability difficult or even impossible. Manufacturers cannot rely on a reference for the implementation and real estate developers and operators are reluctant to commit to a system until it is clear which one will prevail. A similar situation is evident in smart grids and applies equally to the energy supply industry. This fragmentation ultimately leads to missed opportunities in terms of business models which could connect customers with service providers. We present a first draft of an architecture: SINA - Smart Interoperability Architecture. SINA is based on existing decentralized infrastructure, which avoids creating a dependency of the market participants on an overpowering service provider. The core element of the technical solution is an open-source module integrated in the private clouds of the manufacturers, energy suppliers and service providers. The architecture addresses problems of data ownership, privacy and data security avoiding central administrative structures. It manages data access and transfer in a decentralized and distributed system. SINA uses a blockchain and smart contracts to make sure that the pieces of information about which data are accessed, by whom they are accessed, how they are processed, and which monetary transactions take place are immutably stored and made available. This allows providers to offer services to users in a transparent and trustworthy manner. Finally, SINA includes a matchmaking block which helps service providers find potential customers and vice versa. This set of features makes SINA unique.
The Weakly-Compressible Smoothed Particle Hydrodynamics (WCSPH) method is a Lagrangian method that is typically used for the simulation of incompressible fluids. While developing an SPH-based scheme or solver, researchers often verify their code with exact solutions, solutions from other numerical techniques, or experimental data. This typically requires a significant amount of computational effort and does not test the full capabilities of the solver. Furthermore, often this does not yield insights on the convergence of the solver. In this paper we introduce the method of manufactured solutions (MMS) to comprehensively test a WCSPH-based solver in a robust and efficient manner. The MMS is well established in the context of mesh-based numerical solvers. We show how the method can be applied in the context of Lagrangian WCSPH solvers to test the convergence and accuracy of the solver in two and three dimensions, systematically identify any problems with the solver, and test the boundary conditions in an efficient way. We demonstrate this for both a traditional WCSPH scheme as well as for some recently proposed second order convergent WCSPH schemes. Our code is open source and the results of the manuscript are reproducible.
Gurmeet Singh, Anthony M. Waas, Veera Sundararaghavan
Additive manufacturing of a single crystalline metallic column is studied using molecular dynamics simulations. In the model, a melt pool is incrementally added and cooled to a target temperature under isobaric conditions to build a metallic column from bottom up. Common neighbor analysis (CNA) is used to observe the evolution of atomic scale defects during this process. The solidification is seen to proceed along two directions for an added molten layer. The molten layer in contact with the cooler lattice has a fast solidification front that competes with the slower solidification front starting from the top layer. Defect structure formed strongly depends on the speeds of the two competing solidification fronts. Up to a critical layer thickness, defect free single crystals are obtained as the faster solidification front reaches the top of the melt pool before the initiation of the slower front from the top. Slower cooling rates lead to reduction in defects, however, the benefits diminish below a critical rate. Defect content can be significantly reduced by raising the temperature of the powder bed to a critical temperature. This temperature is determined by two competing mechanisms: slower cooling rates at higher temperatures competing against increase in amorphousness as one gets closer to the melting point. Finally, effect of added soft inclusion (SiS2) and a hard inclusion (SiC) on the defect structure is studied. Hard inclusions lead to retained defect structure while soft inclusions reduce defective content compared to a pure metal.
Yunhui Chen, Samuel J. Clark, David M. Collins
et al.
The governing mechanistic behaviour of Directed Energy Deposition Additive Manufacturing (DED-AM) is revealed by a combined in situ and operando synchrotron X-ray imaging and diffraction study of a nickel-base superalloy, IN718. Using a unique process replicator, real-space phase-contrast imaging enables quantification of the melt-pool boundary and flow dynamics during solidification. This imaging knowledge informed precise diffraction measurements of temporally resolved microstructural phases during transformation and stress development with a spatial resolution of 100 $μ$m. The diffraction quantified thermal gradient enabled a dendritic solidification microstructure to be predicted and coupled to the stress orientation and magnitude. The fast cooling rate entirely suppressed the formation of secondary phases or recrystallisation in the solid-state. Upon solidification, the stresses rapidly increase to the yield strength during cooling. This insight, combined with IN718 $'$s large solidification range suggests that the accumulated plasticity exhausts the alloy$'$s ductility, causing liquation cracking. This study has revealed additional fundamental mechanisms governing the formation of highly non-equilibrium microstructures during DED-AM.
Many emerging applications in microscale engineering rely on the fabrication of three-dimensional architectures in inorganic materials. Small-scale additive manufacturing (AM) aspires to provide flexible and facile access to these geometries. Yet, the synthesis of device-grade inorganic materials is still a key challenge towards the implementation of AM in microfabrication. Here, we present a comprehensive overview of the microstructural and mechanical properties of metals fabricated by most state-of-the-art AM methods that offer a spatial resolution $\leq$10$μ$m. Standardized sets of samples were studied by cross-sectional electron microscopy, nanoindentation and microcompression. We show that current microscale AM techniques synthesize metals with a wide range of microstructures and elastic and plastic properties, including materials of dense and crystalline microstructure with excellent mechanical properties that compare well to those of thin-film nanocrystalline materials. The large variation in materials performance can be related to the individual microstructure, which in turn is coupled to the various physico-chemical principles exploited by the different printing methods. The study provides practical guidelines for users of small-scale additive methods and establishes a baseline for the future optimization of the properties of printed metallic objects $-$ a significant step towards the potential establishment of AM techniques in microfabrication.
In a human-centered intelligent manufacturing system, sensing and understanding of the worker's activity are the primary tasks. In this paper, we propose a novel multi-modal approach for worker activity recognition by leveraging information from different sensors and in different modalities. Specifically, a smart armband and a visual camera are applied to capture Inertial Measurement Unit (IMU) signals and videos, respectively. For the IMU signals, we design two novel feature transform mechanisms, in both frequency and spatial domains, to assemble the captured IMU signals as images, which allow using convolutional neural networks to learn the most discriminative features. Along with the above two modalities, we propose two other modalities for the video data, at the video frame and video clip levels, respectively. Each of the four modalities returns a probability distribution on activity prediction. Then, these probability distributions are fused to output the worker activity classification result. A worker activity dataset of 6 activities is established, which at present contains 6 common activities in assembly tasks, i.e., grab a tool/part, hammer a nail, use a power-screwdriver, rest arms, turn a screwdriver, and use a wrench. The developed multi-modal approach is evaluated on this dataset and achieves recognition accuracies as high as 97% and 100% in the leave-one-out and half-half experiments, respectively.
Optics for future X-ray telescopes will be characterized by very large aperture and focal length, and will be made of lightweight materials like glass or silicon in order to keep the total mass within acceptable limits. Optical modules based on thin slumped glass foils are being developed at various institutes, aiming at improving the angular resolution to a few arcsec HEW. Thin mirrors are prone to deform, so they require a careful integration to avoid deformations and even correct forming errors. On the other hand, this offers the opportunity to actively correct the residual deformation: a viable possibility to improve the mirror figure is the application of piezoelectric actuators onto the non-optical side of the mirrors, and several groups are already at work on this approach. The concept we are developing consists of actively integrating thin glass foils with piezoelectric patches, fed by voltages driven by the feedback provided by X-rays. The actuators are commercial components, while the tension signals are carried by a printed circuit obtained by photolithography, and the driving electronic is a multi-channel low power consumption voltage supply developed in-house. Finally, the shape detection and the consequent voltage signal to be provided to the piezoelectric array are determined in X-rays, in intra-focal setup at the XACT facility at INAF/OAPA. In this work, we describe the manufacturing steps to obtain a first active mirror prototype and the very first test performed in X-rays.