InterPore2026

Contact-Aware Grain Mechanics for Improved Elastic and Seismic Property Prediction in Digital Rocks

20 May 2026, 15:35

1h 30m

Poster Presentation (MS09) Pore-Scale Physics and Modeling Poster

Speaker

Erik Glatt (Math2Market GmbH)

Description

Accurate estimation of elastic and seismic properties is a cornerstone of digital rock physics, supporting rock-physics modeling, geomechanics, and reservoir characterization. Reliable numerical prediction of compressional and shear wave velocities (Vp and Vs) is essential for linking pore-scale microstructure to field-scale seismic observations used in reservoir evaluation, well placement, and production monitoring. However, conventional digital rock workflows often systematically overestimate elastic stiffness, primarily due to limited image resolution and simplified representations of grain-contact geometries and mechanics.

This work presents an advanced modeling strategy that addresses this limitation by explicitly incorporating grain-contact mechanics into elastic property estimation. The proposed workflow is applicable across a range of lithologies, including clastic sandstones and carbonate grainstones. High-resolution digital rock images are segmented using a watershed-based approach that enables improved reconstruction of individual grains and their contact networks. Grain-contact areas are explicitly identified, allowing local mechanical properties to be modified based on contact geometry.

Elastic stiffness at grain–grain interfaces is scaled as a function of contact area, accounting for weakening effects that are typically neglected in standard digital rock physics approaches. Using this contact-aware microstructural model, the effective stiffness tensor is computed via numerical homogenization, with the linear elasticity equations solved using an FFT-based framework. Compressional and shear wave velocities are then derived directly from the stiffness tensor and validated against laboratory measurements.

Application of the methodology demonstrates a significant reduction in the overprediction of effective stiffness. Simulated elastic moduli and seismic velocities show close agreement with experimental data, indicating that the approach captures lithology-dependent elastic behavior with improved fidelity. By better representing grain-contact mechanics, the workflow enhances the robustness and accuracy of elastic and seismic property predictions.

The proposed approach strengthens the pore-scale foundation of digital rock physics and enables more reliable scaling from microstructure to seismic response. It provides a practical and broadly applicable framework for improving quantitative seismic interpretation, rock-physics modeling, and reservoir characterization.