Deadline Date: 01 September 2026
The fields of mechanics and materials science are undergoing a profound transformation, driven by the convergence of high-fidelity computational methods and groundbreaking advances in artificial intelligence (AI). Accurately describing the mechanical behavior of materials is of great significance for their engineering applications. Traditional computational techniques, such as the Finite Element Method (FEM), Meshfree Methods (MM), Discrete Element Method (DEM), and Molecular Dynamics (MD), have long been the cornerstone for simulating complex physical phenomena. However, they often face significant challenges in dealing with multi-scale problems, inverse design, and extreme computational costs.
With the rapid development of artificial intelligence (AI), advanced techniques such as neural networks are expected to bridge the gap in mechanical behavior in different scales, driving traditional material mechanics research from a "physics-driven" paradigm to a "data-physics hybrid-driven" paradigm. AI technologies not only accelerate the modeling and simulation of complex systems but also show great potential in uncovering underlying physical laws, predicting mechanical behaviors, and optimizing material structures. In particular, methods such as neural networks and graph learning provide novel approaches to address challenges in multi-physics coupling, spatiotemporal predictions, and nonlinear behavior characterization.
Despite the widespread application of neural network technologies in material mechanics, numerous challenges remain. Many data-driven computational methods lack explicit constraints on physical laws such as constitutive relationships and energy conservation, leading to predictions that violate fundamental mechanical principles. Additionally, these methods suffer from strong data dependency, high costs of obtaining high-quality material data, poor interpretability, and limited generalization capabilities. Although physics-informed neural networks (PINNs) partially alleviate these issues by incorporating physical constraints, they still face challenges such as limited adaptability to complex scenarios and the absence of robust theoretical validation frameworks. These technical bottlenecks highlight the urgent need for breakthroughs in both theory and practice within the field of intelligent computation for material mechanics.
This special issue aims to bring together cutting-edge research at the intersection of AI, computational mechanics, and materials science, addressing the core challenges of the current "data-physics hybrid-driven" paradigm. We encourage original research that overcomes the limitations of traditional computational mechanics models, enhances physical constraints, and improves the generalization and interpretability of AI models. Priority topics include:
Exploring methods to enhance model transparency and interpretability, and establishing a quantitative evaluation framework for computational mechanics and AI-based mechanics models.
Addressing the scarcity of high-quality material data by investigating small-sample learning methods and data augmentation strategies in AI mechanics.
Leveraging generative AI technologies to design novel materials with specific microstructures, molecular configurations, or macroscopic structures based on target performance requirements, enabling AI-driven inverse material design.
Integrating traditional physical simulations (e.g., finite element methods, meshfree methods) with AI surrogate models to construct efficient hybrid analysis frameworks for accelerating the solution of complex nonlinear dynamics problems in materials.
We invite submissions from theoretical, computational, and applied research in material mechanics, computational science, and artificial intelligence, particularly those demonstrating methodological rigor, experimental validation, and engineering applicability. Topics of interest include, but are not limited to:
Advanced Computational Methods: Material mechanics, Multi-scale and multi-physics modeling (such as FEM, MM, DEM, MD, etc.), phase-field modeling and fracture mechanics, contact mechanics and large deformation analysis, as well as uncertainty quantification and probabilistic modeling, and so on.
Physics-informed AI algorithms: Advances in physics-informed neural networks for material mechanics and other methods embedding physical constraints.
Automated scientific discovery: Applications of symbolic regression, tensor networks, and other methods for automatically discovering or identifying material constitutive relationships and physical laws.
Intelligent microstructure characterization and generation: Graph neural networks for microstructure characterization and damage evolution, and generative models for virtual microstructure reconstruction and optimization.
AI-enabled multiscale modeling: Constructing efficient surrogate models to replace computationally expensive microscale simulations by learning from high-fidelity simulation data and coupling information across scales.
Intelligent multiphysics coupling: AI solvers and hybrid modeling strategies for complex coupled problems such as thermo-mechanical-chemical-electrical interactions.
Data-driven mechanics models: Data-driven paradigms for crystal plasticity, phase-field models, and damage mechanics.
AI technologies for additive manufacturing: Real-time simulation of 3D printing processes, defect prediction, and AI-based optimization of process parameters.
Intelligent performance evaluation and health monitoring: AI-based predictions of material fatigue life, fracture behavior analysis, and intelligent evaluation of structural health.
Inverse design of new structures and materials: Searching for optimal structures or material compositions to achieve given performance targets, such as odd elasticity theory, functionally graded materials, plate and shells, metamaterials mechanics.
AI-driven digital twins: Developing AI surrogate models for real-time prediction, decision support, and performance optimization in material and structural systems.
This Special Issue aims to gather these frontier studies to provide powerful computational tools for addressing major challenges in advanced materials, accelerating the innovation cycle from material mechanics to engineering application.
Editorial board
