PhD position: AI-driven surrogate approaches for microstructure-aware structural modeling University of Lorraine in France

This PhD position is part of the AMMETIS project (AI-assisted Simulations of Microstructure driven Mechanical properties from high Throughput and multiscale analysis), within the PEPR DIADEM framework. The project aims to develop an advanced platform for characterizing innovative materials by integrating experimental techniques, mesoscopic modeling, and artificial intelligence. High-throughput experiments and large-scale numerical simulations will generate rich datasets describing the relationship between microstructure, deformation mechanisms, and mechanical response.

Physics-based simulations using advanced mesoscopic crystal plasticity offer powerful predictive capabilities but are computationally expensive for realistic microstructures and large-scale analyses. The main challenge addressed in this PhD is to develop efficient AI-driven surrogate models that can rapidly predict macroscopic mechanical properties from microstructural descriptors, while preserving the underlying physical mechanisms.

The PhD project will leverage large datasets generated within AMMETIS, combining high-resolution experiments (HR-DIC, HR-EBSD, nanoindentation mapping) and numerical simulations using FFT-based crystal plasticity platforms. Various machine learning strategies will be explored, including deep learning architectures for microstructure-property mapping, convolutional neural networks for image analysis, graph-based representations, physics-oriented descriptor discovery (RRAE), and neural operator approaches for approximating complex mechanical solutions. Special emphasis will be placed on integrating physics-informed constraints to ensure robustness, interpretability, and extrapolation capabilities.

The resulting surrogate models will enable fast prediction of effective mechanical properties and deformation fields for complex microstructures, bridging mesoscale simulations and structural-scale applications. These tools will accelerate the exploration of microstructure-property relationships and open new perspectives for the design and optimization of advanced structural materials.

The research will be conducted primarily at PIMM (Laboratoire Procédés et Ingénierie en Mécanique et Matériaux), Paris, in collaboration with LEM3 (Laboratoire d’Études des Microstructures et de Mécanique des Matériaux), University of Lorraine, Metz. The position offers a gross salary of approximately €2300 per month for three years, starting in September 2026 (flexible date). The working language is English, and the position is not funded by an EU programme.

Applicants should have a Master’s degree in Mechanical Engineering, Materials Science, Applied Mathematics, Data Science, or Computational Mechanics, with strong skills in continuum mechanics, numerical modeling, machine learning, and scientific computing. Experience with Python, PyTorch, TensorFlow, and numerical methods for PDEs is required. Candidates should be motivated for collaborative academic-industrial research and able to work in interdisciplinary environments.

To apply, send your application and supporting documents by email to [email protected], referencing the position title. Attach your CV, cover letter, and transcripts. Download and complete the internal application form if required. The application deadline is 30 June 2026.