Bounding the Multi-Scale Domain in Numerical Modelling and Meta-Heuristics Optimization: Application to Poroelastic...

Area of Science:

• Continuum Mechanics
• Materials Science
• Computational Mechanics

Background:

• Materials often exhibit periodic or quasi-periodic micro-structures influencing macroscopic behavior under load.
• Characterizing micro-scale behavior experimentally is challenging, necessitating advanced modeling techniques.
• Poroelastic media with damage and cracks present complex mechanical responses at the micro-scale.

Purpose of the Study:

• To develop and validate a multi-scale constitutive law for cracked, damageable, poroelastic media.
• To employ meta-heuristics for calibrating micro-scale parameters essential for accurate macroscopic predictions.
• To demonstrate the approach's efficacy in modeling the behavior of micro-fractured geological materials.

Main Methods:

• Asymptotic homogenization technique applied to a cracked, damageable, poroelastic medium.
• Development of a multi-scale constitutive law for predicting macroscopic material behavior.
• Utilization of meta-heuristics for calibrating micro-scale parameters against a synthetic failure envelope.

Main Results:

• The proposed multi-scale approach effectively models the macroscopic behavior of micro-fractured materials.
• Calibration using meta-heuristics overcomes limitations in experimental micro-scale data.
• The constitutive law accurately predicts the failure envelope for materials like coal and crystalline rocks.

Conclusions:

• The validated multi-scale constitutive law provides a robust framework for analyzing complex material behaviors.
• Meta-heuristics offer a powerful tool for parameter calibration in multi-scale material modeling.
• This approach is particularly effective for simulating the mechanical response of naturally fractured rocks and similar materials.