A Two-Stage Reconstruction of Microstructures with Arbitrarily Shaped Inclusions

Area of Science:

• Materials Science
• Computational Modeling
• Statistical Physics

Background:

• Statistical reconstruction of heterogeneous microstructures is crucial for material design and analysis.
• Existing entropy methods for reconstruction have limitations in handling complex inclusion shapes and scales.

Purpose of the Study:

• To develop an effective and widely applicable method for the statistical reconstruction of heterogeneous microstructures with compact inclusions of any shape.
• To introduce multi-scale extended entropic descriptors (ED) for quantifying spatial non-uniformity in object configurations.

Main Methods:

• A two-stage reconstruction (TSR) process is employed.
• Stage 1: Generation of synthetic clusters with target areas and interfaces but random shapes, followed by selection based on a cost function using extended ED.
• Stage 2: Modification of simulated annealing (SA) by randomly moving synthetic clusters instead of swapping pixels.

Main Results:

• The TSR method successfully reconstructs complex two-phase microstructures, including irregular silica inclusions in rubber and stones in cement paste.
• The method demonstrates convincing statistical mapping of inclusion shapes.
• Validation through reconstruction of specific complex microstructures confirms the method's accuracy.

Conclusions:

• The two-stage reconstruction (TSR) method offers an effective solution for reconstructing heterogeneous microstructures with complex inclusions.
• Key advantages include ease of generating synthetic microstructures, low computational cost, and satisfactory statistical accuracy.
• The method's simplicity facilitates independent applications in various scientific and engineering fields.