Abstract In the context of structural health monitoring (SHM), the selection and extraction of damage-sensitive features from raw sensor recordings represent a critical step towards solving the inverse problem underlying the identification of structural health conditions. This work introduces a novel approach that employs deep neural networks to enhance stochastic SHM methods. A learnable feature extractor and a feature-oriented surrogate model are synergistically exploited to evaluate a likelihood function within a Markov chain Monte Carlo sampling algorithm. The feature extractor undergoes pairwise supervised training to map sensor recordings onto a low-dimensional metric space, which encapsulates the sensitivity to structural health parameters. The surrogate model maps structural health parameters to their feature representation. The procedure enables the updating of beliefs about structural health parameters, eliminating the need for computationally expensive numerical models. A preliminary offline phase involves the generation of a labeled dataset to train both the feature extractor and the surrogate model. Within a simulation-based SHM framework, training vibration responses are efficiently generated using a multi-fidelity surrogate modeling strategy to approximate sensor recordings under varying damage and operational conditions. The multi-fidelity surrogate exploits model order reduction and artificial neural networks to speed up the data generation phase while ensuring the damage-sensitivity of the approximated signals. The proposed strategy is assessed through three synthetic case studies, demonstrating high accuracy in the estimated parameters and strong computational efficiency.

Abstract Effective battery thermal management system (BTMS) is critical for lithium-ion battery (LIB) safety and performance in electric vehicles. This study presents a CFD-driven optimisation framework for an immersion cooling BTMS using sustainable palm biodiesel as coolant. The Multi-scale Multi-Domain (NTGK) framework is conducted to effectively capture the complex interactions among various physicochemical processes. The Electrochemical-thermal Model (ECM) is applied using the Newman, Tiedeman, Gu, and Kim (NTGK) model. A conjugate heat transfer model for a 3S2P pouch cell module (20 Ah LiFePO₄) is developed and validated against experimental data (< 2% error). The CFD model of a battery module is developed to train an ultra-fast metamodel for battery geometry optimisation. Two key parameters are optimised, namely: battery gap spacing (3–10 mm) and inlet/outlet width (5–15 mm), via Optimal Latin Hypercube Sampling, Support Vector Regression, and GDE3 algorithm. Palm biodiesel is used as a dielectric coolant in the proposed system to preserve LIB temperature within 20–40 $$^\circ{\rm C} $$ , preventing thermal runaway and ensuring a lightweight BTMS design. Compared to a conventional 3M-Novec, the palm biodiesel achieved system lightweight by 43%. The findings can establish biofuel immersion cooling as an eco-friendly BTMS solution, achieving Pareto-optimal figures: T max < 29.9°C, Δ T < 5°C, and Δ P < 145.275 Pa (at 5C and 0.05 m/s).

Abstract We introduce , an open-source Python library designed for mesh agglomeration in both two- and three-dimensions, based on employing Graph Neural Networks (GNN). serves as a comprehensive solution for training a variety of GNN models, integrating deep learning and other advanced algorithms such as METIS and k-means to facilitate mesh agglomeration and quality metric computation. The library’s introduction is outlined through its code structure and primary features. The GNN framework adopts a graph bisection methodology that capitalizes on connectivity and geometric mesh information via SAGE convolutional layers, in line with the methodology proposed in (Antonietti and Manuzzi in J Comput Phys 452:110900, 2022; Antonietti et al. in Polytopal mesh agglomeration via geometrical deep learning for three-dimensional heterogeneous domains, arXiv:2406.10587 , 2024). Additionally, the proposed library incorporates reinforcement learning to enhance the accuracy and robustness of the model initially suggested in [1, 2] for predicting coarse partitions within a multilevel framework. A detailed tutorial is provided to guide the user through the process of mesh agglomeration and the training of a GNN bisection model. We present several examples of mesh agglomeration conducted by , demonstrating the library’s applicability across various scenarios. Furthermore, the performance of the newly introduced models is contrasted with that of METIS and k-means, illustrating that the proposed GNN models are competitive regarding partition quality and computational efficiency. Finally, we exhibit the versatility of ’s interface through its integration with , an open-source library implementing discontinuous Galerkin methods on polytopal grids for the numerical discretization of multiphysics differential problems.

Abstract This paper presents two approaches: the virtual element method (VEM) and the stabilization-free virtual element method (SFVEM) for analyzing thermomechanical behavior in electronic packaging structures with geometric multi-scale features. Since the virtual element method allows the use of arbitrary polygonal elements, the inherent mesh flexibility of VEM allows localized mesh modifications without affecting global mesh structure, making it particularly effective for the analysis of electronic packaging reliability involving complex geometries and multiple geometric scales. The approach implements a novel non-matching mesh generation strategy that strategically combines polygonal meshes for complex small-scale regions with regular quadrilateral meshes for larger domains. The VEM formulation addresses both heat conduction and thermomechanical coupling problems, with comprehensive verification through analytical benchmarks and practical electronic packaging case studies, including Through-Silicon Via (TSV), Ball Grid Array (BGA), and Plastic Ball Grid Array (PBGA) structures. Results demonstrate that the method accurately captures stress concentrations at material interfaces and provides reliable thermal and mechanical response predictions. Some MATLAB codes for the numerical examples are provided at https://github.com/yanpeng-gong/VEM-electronic-packaging and on the VEMhub website ( www.vemhub.com ).