With the unceasing development of both computational capabilities and advanced experimental setups (e.g. tomography devices), researchers and engineers are now facing the issues of identification, representation and simulation of material behavior at unprecedented levels of resolution. Such developments advocate for the construction of robust multimodel and multiscale predictors accommodating randomness and uncertainties in a high dimensional setting. This symposium aims at bringing together researchers involved in the development of stochastic methods and algorithms for the multiscale analysis of engineered or natural heterogeneous materials.
Contributions to the following topics are specifically encouraged:
• Algebraic, functional and morphological representations in high dimensional spaces.
• Concurrent or sequential coupling of stochastic models defined at different scales.
• Simulation algorithms (for stochastic processes, random fields and random sets) and their application to multiscale analysis.
• Statistical inverse identification of multiscale systems.
• Stochastic (space/time) homogenization and related numerical methods.