In this work, a novel Thin Layer Compression testing rig by the company ZwickRoell was used to characterize the compression behavior of individual sheets of electrode foils from a commercial Li-ion battery cell. These electrode foils consist of a metallic current collector layer sandwiched between two active material coating layers. A custom mathematical correction method was developed, facilitated by the Thin Layer Compression testing rig, to separate the obtained data into the responses of the individual foil layers. This method enabled extraction of the stress–strain response of only the active material coating without the necessity of changing the structure of the electrode foils for the measurements. Finite element models replicating the sandwich structure of the electrode foils were created for the commercial solvers LS-DYNA and ABAQUS, where separate material models were implemented for current collector layers and active material coating layers based on the compression experiments and data from the literature. The force–deformation results of the simulations in LS-DYNA and ABAQUS show good agreement with the experimentally obtained data, despite the significant differences of how material models and elements are implemented in the two finite element method solvers.

Warpage prediction and control in package-on-package assemblies require efficient metamodels capable of handling high-dimensional thermo-mechanical design spaces. This work develops and compares five metamodels: artificial neural network (ANN), Gaussian process regressor (GPR), Transformer, Transformer-ANN hybrid, and Transformer-GPR hybrid, trained on 6000 finite-element simulations to predict the full-field top-surface deformation from 143 material, geometric, and process-temperature features. The Transformer-GPR hybrid achieved the highest predictive accuracy and was therefore employed in a constrained particle swarm optimization (PSO) workflow, demonstrating substantial warpage reduction across representative package configurations. Transformer-based attention maps, together with one-at-a-time perturbation analyses, identified layer thickness and coefficient of thermal expansion (CTE) as dominant drivers of warpage. Moreover, outer routing-layer CTEs and thicknesses exhibited the strongest influence within their respective classes, enabling identification of small, high-impact feature subsets for manufacturable design optimization. These findings highlight the potential of Transformer-enhanced metamodels not only for accurate warpage prediction but also for guiding reduced-variable optimization in practical package-design workflows.

Printed circuit board assemblies (PCBAs) in surgical medical devices require encapsulation with thermoset moulding compounds to withstand chemical cleaning and steam sterilisation. The thermoset curing process, however, generates high stresses in solder joints of surface-mounted devices that can cause damage. This paper presents a virtual prototyping process for PCBA encapsulation using the open-source tools GOMA and code_aster. The process covers the filling, packing, and cooldown phases, incorporating cure-degree-, temperature-, and shear-rate-dependent viscosity via the Cross–Castro–Macosko model and cure-dependent mechanical properties via the Cure Hardening Instantaneous Linear Elastic (CHILE) model with incremental stress calculation. Combined with the ANAND viscoplastic model for lead-free solder, inelastic solder-joint strains are computed in a nonlinear thermo-mechanical analysis. Validation against a commercial solver confirms that the PCBA encapsulation can be accurately modelled with open-source tools.

Thermoset molding processes are widely used in the manufacturing of printed circuit boards, power electronics encapsulation, and high-performance composite structures. These processes involve strongly coupled thermal, rheological, and chemical phenomena that directly influence mold filling behavior, curing kinetics, and thermally induced stresses. Although high-fidelity multiphysics simulations based on Computational Fluid Dynamics (CFD) and Finite Element Analysis (FEA) provide detailed predictions of these interactions, their high computational cost limits their applicability for design exploration and real-time process monitoring.This work presents a surrogate-integrated, simulation-driven digital twin for thermoset molding processes that combines multiphysics simulations with data-driven surrogate modeling. Long Short-Term Memory (LSTM) networks are trained to learn the temporal evolution of temperature. The resulting surrogate model predicts thermo-chemical field evolution within milliseconds while maintaining strong agreement with reference simulations. The proposed framework enables fast predictive capabilities suitable for realtime digital twins and data-driven process optimization in thermoset manufacturing.

Femtosecond laser micromachining in transparent FR4 epoxy substrates was performed using a 515 nm green laser system, with nanometer-scale surface topography characterized via white light interferometry (WLI). A Super-Gaussian fitting methodology was applied to quantify ablation crater profiles, introducing a shape exponent parameter that simultaneously describes bottom flatness and sidewall steepness characteristics. This approach achieved 59.9 % reduction in normalized root-mean-square error (NRMSE) compared to conventional parabolic fitting techniques. Through systematic full factorial experimental design spanning laser powers from 3 to 18 W and pulse counts from 5 to 200, three semi-empirical predictive models were developed correlating laser processing parameters to Super-Gaussian geometric descriptors: depth (D), diameter (W), and shape exponent (p). These parameters have been predicted by the models with great accuracy, enabling sub-micron geometric control. This investigation establishes, for the first time, a comprehensive physics-informed modeling framework that enables predictive process control in femtosecond laser microfabrication, providing quantitative foundations for automated manufacturing systems and advancing industrial implementation of ultrafast laser technologies in precision polymer composite processing.

Metamaterials with their distinctive unit cell-based periodic architecture feature a wide range of possible properties with unusual characteristics and a high potential for optimization. Due to their complex interaction between unit cell geometry and material properties, as well as their inherent multi-scale nature, suitable optimization strategies need to be developed for metamaterials. One potential approach is to optimize the distribution of unit cells within a part to achieve a predefined deformation response. However, a significant challenge lies in determining the appropriate number and distribution of areas with varying properties (material sections) to facilitate an efficient optimization. In this study, a variable material section discretization scheme is presented, which is aimed at automatically updating the discretization to enhance the efficiency of metamaterial optimizations. This scheme is implemented as an extension to a previously proposed Finite Element simulation-based optimization framework for unit cell-based metamaterials. The framework includes a numerical homogenization method and interpolation scheme for efficiently correlating unit cell parameters with homogenized material properties, coupled with a black-box optimization method. In the present study, the previously proposed framework was extended to incorporate a scheme for monitoring and adjusting the material section discretization during the optimization process. To assess the effectiveness of the implemented routine, it was tested in conjunction with a genetic algorithm for optimizing the parameter distribution of a 2D tri-anti-chiral metamaterial to match a predefined lateral deformation under load.

This study presents an innovative method for integrating Moldflow, Abaqus, and Python to analyze process-related residual stresses in medical PCB encapsulations manufactured via injection molding. The proposed approach enables improved pre-processing in Abaqus, independent of the Moldflow mesh, facilitating the application of boundary conditions and local mesh refinements. The investigation reveals that the curing of the thermoset in the cavity during the injection molding cycle leads to very high temperatures due to the exothermic curing reaction. Process-related residual stresses primarily form during the cooling phase and reach their maximum at the end of this phase when the part has fully cooled to room temperature. Various adaptations of the CHILE (Cure Hardening Instantaneous Linear Elastic) model are presented and compared, resulting in slightly different outcomes regarding the magnitude of process-induced residual stresses. These findings provide valuable insights for optimizing encapsulation processes, not only in medical technology, and contribute to improving the reliability and lifespan of medical PCB assemblies.

Managing substrate warpage is critical in the microelectronics industry, as excessive deformation can lead to strain accumulation in packages, when integrated with components such as ASICs, emitters, and detectors. While finite element analysis is traditionally used for material optimization, the complex structure of laminate substrates makes this approach computationally intensive. This study explores the use of machine learning-based metamodels to accelerate material optimization of laminate substrates. Artificial neural networks and Gaussian process regression models were developed to predict top-surface deformation during reflow processes. In a two-phase approach, Phase A uses the coefficient of thermal expansion as input, while Phase B incorporates coefficient of thermal expansion, elastic modulus, and Poisson's ratio. The best-performing metamodel in each phase is integrated with particle swarm optimization to identify optimal material properties. To our knowledge, this is the first study to jointly optimize coefficient of thermal expansion, elastic modulus, and Poisson's ratio using metamodels to achieve reduced substrate warpage. The proposed method not only reduces computational cost but also achieves higher accuracy than previously reported approaches, demonstrating its effectiveness in minimizing substrate warpage.

This review synthesizes recent progress in femtosecond (fs) laser drilling of epoxy composites, focusing on three interconnected themes critical for industrial scalability: ablation threshold dynamics, multiscale modeling, and adaptive process control. It quantifies how incubation effects reduce multi-pulse ablation thresholds, enabling precise material removal with minimal heat-affected zones. A unified multiscale simulation framework is presented, integrating continuum models, molecular dynamics, and plasma simulations to capture ultrafast energy deposition and stress evolution across vast spatiotemporal scales. To overcome throughput limitations, advanced process control strategies leveraging multimodal diagnostics and machine learning are discussed, demonstrating sub-micron precision and enhanced efficiency. The review concludes by identifying key challenges, such as the need for standardized characterization methods and low-latency feedback systems, and outlines a roadmap for future research, emphasizing the potential for femtosecond laser drilling in high-precision manufacturing applications.

The pursuit of smaller, lighter, and more efficient electronic devices in the electronics industry necessitates advanced packaging solutions. This study investigates the nuanced aspects of microvia technologies. Focusing on the impact of geometrical design parameters on the thermomechanical performance of microvias, Finite Element Method (FEM) simulations provide insights for various combinations of the geometrical variations.The research underscores the critical role of lower microvia wall angles experience failures, corroborating literature predictions. Stress distribution analyses in global and submodel configurations offer consistent trends, emphasizing the importance of meticulous microvia design considerations. The observed strain values and localized failure regions further highlight the need for precision in design to mitigate vulnerabilities. This study contributes essential knowledge to ongoing efforts in advancing microelectronic reliability, sustainability, and innovation, setting the stage for further scientific increment in microelectronic industry.

Laser drilling has gained widespread popularity in industrial applications due to its precision, non-contact nature, speed, efficiency, and energy-saving attributes. This technology plays a vital role in the manufacturing of printed circuit boards (PCBs) and semiconductors. Additionally, femtosecond laser drilling has enhanced the capabilities of micro- and nano-fabrication of materials. The reliability of PCBs is heavily influenced by the laser drilling process. During laser drilling, numerous physical and chemical changes occur within a short timeframe, and the complexity of these changes remains a subject of ongoing investigation. Given the challenges of observing such rapid material transformations through experiments, simulation, and computational analysis offer a cost-effective alternative. KratosMultiphysics is a multi-physics solver employed to address a variety of process simulations across different fields. This paper initially gathered material parameters and recorded various test results under different laser drilling conditions for comparison with simulation outcomes. Utilizing KratosMultiphysics as a computational tool, we simulate the ablation of material with the subsequent heat conduction and decomposition of the polymer material according to a new numerical model.

This review analyzes the state-of-the-art in metamodel-based simulation and optimization of microelectronics package substrate warpage. The work gives an overview on the general procedure, including metamodel selection, Design of Experiment approaches, or sampling. It attempts to not only focus on material property optimization, but also highlight approaches in structural and process optimization, which are largely underexplored. Additionally, the review reveals a critical gap, as no research explicitly incorporates geometry, boundary conditions and constraints as parameters within the metamodel itself. It also summarizes literature on neural network approaches and shows that although there is a significant influence of feature position on substrate warpage, e.g. convolutional neural networks are not well represented in current research.

Mechanical metamaterials have gained a lot of research interest over the last years due to their unusual mechanical properties and potential use for structural applications. However, the design and analysis of mechanical metamaterials remains challenging and time-consuming. Herein, we present a software framework for automated generation, finite element modeling and analysis of extruded mechanical metamaterials based on simple closed curves. By generalizing extruded unit cells with pores defined by simple closed curves, a wide variety of existing and novel metamaterials can be created and analyzed. Each pore can either be empty or filled with one or more materials, resulting in single- or multi-material metamaterials. Since part of the mechanical response of a metamaterial is defined by the geometric parameters of a unit cell, parameter studies are directly integrated into the framework. Examples of well-established mechanical metamaterials were used as benchmark structures and compared to their analytical solutions. We also demonstrate how generalized curves can further improve the mechanical properties of these structures, for example the load bearing capabilities and range of Poisson's ratios. Furthermore, newly developed single- and multi-material designs of mechanical metamaterials with tunable Poisson's ratio are presented and analyzed with the proposed framework. The framework is implemented in Python, executed via ABAQUS and allows for unit-cell based homogenization and full-size 2D and 3D simulations. The scripts are made open source and publicly available.

Reliability testing is essential in the PCBA (Printed Circuit Board Assembly) manufacturing process since it helps to identify potential issues before they become significant in the future. This shall ensure the quality and long-term performance of the product. However, the challenge is that the experimental reliability test during manufacturing is quite time-consuming and cost-consuming, and finding a new method to accelerate the test itself is very difficult. Currently, virtual performance predictors such as the finite element method (FEM) form a key for assessing the criticality of a PCBA product, but they are still limited due to their costly computational power if the problem is geometrically complex or difficult to handle, as electronic devices become more compact and sophisticated. With the rapid development of artificial intelligence, novel modelling techniques involving machine learning (ML) and deep learning (DL) have gained extensive attention from engineers and this is especially the case in the field of electronic system reliability. By building an appropriate surrogate model, designers need to input only the component geometry of the newly designed electronic product to immediately obtain a lifetime estimation of the specific product. The given work reviews the state-of-the-art surrogate model-based approaches used in electronic systems, specifically in the field of PCBA. The challenges and future development of surrogate modelling in support of global criticality assessment at the PCBA level are also analysed and discussed.

Among several new module concepts and cell architectures, shingled solar cell interconnection is a promising technology to realize increased power output (higher power densities) by increased active cell area and low electrical losses. The direct interconnection by a slight overlapping of solar cells, using electrically conductive adhesive (ECA) asjoint material, requires a sound understanding of the loads the ECA joint faces during module lifetime. The ECA joint not only has to enable electric conductivity but at the same time needs to compensate different mechanical responses of the materials in the module laminate due to varying external stresses induced by variations of temperature and mechanical loads deflecting the module, e.g. wind or snow covering the modules. Especially thermal effects on the structure of solar panels are significant and complicated due to differing coefficients of thermal expansion (CTE) of the materials in use. However, the influence of anisotropic and also possibly negative coefficients of thermal expansion on thermomechanical stresses in string of shingled solar cells has not yet been investigated. Therefore, the current work aims for a more realistic representation of the thermomechanical behavior of photovoltaic (PV) modules by considering the anisotropy of the CTE of polymer films.

Warpage is an issue in the manufacturing of electronic packages. The main driver for this effect is the mismatch of material properties especially during temperature cycling. To predict warpage before building samples, the FEM (Finite Element Method) is used in this study. PCBs (Printed Circuit Boards) are assembled in array format. Thus the simulation should be done in this format. Studies investigated PCBs thru a detailed representation [1]. However, no studies were found which simulate an array of PCBAs (Assembled PCBs). This is mainly due limitation of the hardware memory. The array of m2x modules in this work is not feasible to be calculation thru a usual modelling approach, even not through cloud computing. Thus this study investigates the use of substructures to overcome the hardware limitations. The array was manufactured and the simulation results were validated by using DIC (Digital Image Correlation). The predicted warpage of the novel modelling approach is well aligned with the measurement results.The establishment of the method for array format warpage assessment significantly decreases the development times for electronic packages, due to the mitigation of reliability risks and thus design cycles.

Deriving the effective material properties of printed circuit board (PCB) layers is a particular challenge due to the intricate patterns of the conductive artwork. In this paper, we develop a convolutional neural network model to predict the effective orthotropic thermo-mechanical properties of the copper conductive layers of PCBs. To ensure the necessary level of detail for the material modeling, the images of the copper-resin distributions are extracted from the real electronic design files as patches of preselected pixel sizes. For each image, the corresponding homogenized material properties are computed using finite element analyses. After the training process on the formed dataset is completed, the developed model is validated using data unseen during training. The obtained results indicate that the considered approach is suitable for the prediction of the effective orthotropic properties of copper-resin patterns, and moreover, is promising for facilitating present-day time-to-market objectives in the industry.

Different designs for mechanical metamaterials with tunable normal-strain shear coupling effect have been demonstrated over the last years. Their adjustable shear deformation makes them suitable as building blocks for soft robotics applications or structures with a desired deformation behavior, such as shape morphing structures. Herein, we present a modified 2.5D and 3D chiral-based mechanical metamaterials with tunable normal-strain shear coupling effect and Poisson's Ratios close to zero. Advancing from conventional chiral-based metamaterials by introducing additional geometric freedoms into the design of the unit cell, a broad range of shear deformations, compression moduli and porosities can be achieved. 2.5D specimens with selected geometric parameters were additively manufactured with polypropylene using Fused Filament Fabrication. Compression tests were performed to investigate the mechanical properties and shear deformation. Two different numerical models were employed using ABAQUS to study the influence of the geometric parameters onto the mechanical properties and were verified by the experiments. The numerical material models were based on three-point-bending test data. The three-dimensional design was investigated with numerical simulations based on a homogenization approach to cover a broad range of geometric parameters.

Due to the steady development of conductive filled resins and multi-material techniques for the digital light processing (DLP) additive manufacturing technology, fabrication of complex conductive structures for microelectronic applications is becoming a potential use case for this technology. When processing electrically conductive systems, temperature effects are of special importance, as the highly filled systems often need elevated temperatures for an increase in reaction rate and to decrease viscosity to achieve printability. Thus, an accurate calculation of the temperature distribution during the process is needed for accurate process modeling. This study describes the thermomechanical material characterization and kinetic modelling including exothermic heat generation during curing for a thermal simulation as part of a DLP process simulation framework. The key properties for the simulation, such as specific heat capacity, thermal diffusivity and reaction enthalpy were characterized and compared between specifically developed conductive and non-conductive acrylic resins. In addition, the temperature dependent mechanical properties were studied and the electrical conductivity of the filled material was measured. A simulation was set up to calculate the heat generation due to the exothermic reaction during printing and the results were validated against measured data from printing trials with the nonconductive material. The results of this work show the temperature dependence of important properties of unfilled and electrically conductive materials for the thermal DLP process simulation and the capabilities of the proposed simulation strategy to calculate the temperature distribution during the process.

Metamaterials are a class of materials with a distinctive unit cell-based periodic architecture, often resulting in unique mechanical properties. The potential of metamaterials can be further improved by using gradients of unit cell parameters and thereby creating a specific distribution of material properties in a part. This design freedom comes with the challenge of finding new suitable design and optimization strategies. In this study, a Finite Element simulation-based optimization framework to create a predefined deformation behavior of unit cell-based metamaterials is presented. The framework consists of a numerical homogenization method to create an efficient linear elastic material representation, an interpolation scheme for a fast correlation of unit cell parameters with homogenized material properties and a black-box based optimization part. The framework is tested on a tri-anti-chiral metamaterial with additional geometric parameters and a newly developed transition unit cell. Several specified lateral deformations of simple 2D rectangular test specimens under tensile load are used as primary optimization targets and the homogenized results are validated against Finite Element simulations of fully modeled tri-anti-chiral structures. In addition, the effect of the design space and the number of design variables is studied and several combinations of optimization objectives and constraints are tested.