The article considers an approach for multiscale geomechanical modeling under finite strains. A mathematical model, methods and algorithms for the multiscale numerical estimation of the effective elastic and thermal properties of the preloaded heterogeneous porous materials under finite strains are presented. The developed algorithms were applied to the problem of the estimation of the effective properties of core samples. The digital models obtained from computed tomography scan data of core samples are used. An initial voxel representation of the digital core sample is transformed into an unstructured mesh, thus reducing the number of unknowns by orders of magnitude and requiring computational resources accordingly. The mesh convergence tests demonstrated the efficiency and correctness of the developed algorithms. The calculations of the effective elastic and thermal properties of preloaded porous materials are performed with CAE Fidesys using finite element method. The numerical results demonstrate the significant impact of pre-loading by internal pressure on the effective properties of porous materials.

本文考虑了一种在有限应变下进行多尺度地质力学建模的方法。提出了一种数学模型、方法和算法，用于在有限应变下对预加载的异质多孔材料的有效弹性和热性能进行多尺度数值估计。所开发的算法被应用于估计岩心样品的有效性质的问题。使用从岩心样品的计算机断层扫描数据获得的数字模型。数字岩心样本的初始体素表示被转换为非结构化网格，从而将未知数的数量减少了几个数量级，并相应地需要计算资源。网格收敛测试证明了所开发算法的效率和正确性。使用 CAE Fidesys 使用有限元方法计算预加载多孔材料的有效弹性和热性能。数值结果证明了内压预加载对多孔材料有效性能的显着影响。