关键词: 岩石力学, 孔洞, 裂隙, 数字图像相关方法, 有效方差, 裂纹识别

Abstract: In order to study the mechanical properties and failure mechanisms of the rocks with groups of holes and flaws, we use the 3D sand printing technique to prepare the rock-like specimens that each contains one hole and two flaws. We use the digital image correlation (DIC) method to monitor the full field of the specimens during their compression process without any contact. By calculating the matrix of the horizontal, vertical and shear strain, we introduce the effective variance of the strain field, which can be used to quantify and identify the failure behavior of the specimens. The main results are listed as below. The mechanical properties of the standard 3D sand printed specimens are similar to those of natural sandstones, and the variations of their mechanical properties are considerably low during the experiments. Therefore, they can be grouped as a rock-like material. Due to the inclusion of the flaws, the mechanical properties of the specimens are degraded. The compressive strength and elastic modulus are reduced by 8.04%-38.91% and 14.44%-27.78%, respectively, compared with those of the specimens containing no flaw but only one hole. Based on the results of DIC, three basic types of cracks are identified successfully, that is, tensile crack (Mode I), shear crack (Mode II) and tensile and shear mixed crack (Mode I-II). The failure patterns of all the specimens with one hole and two flaws show tensile and shear mixed crack (Model I-II). The coalescence patterns between the hole and the flaws can be influenced by their horizontal distance. Those patterns can be classified into tensile coalescence, rotation coalescence and shear coalescence. The discreteness of the strain field can be quantified by the effective variance of strain field comprehensively. The effective variance of the strain field is close to zero at the initial crack closure stage and the elastic deformation stage. It grows differently after the appearance of cracks. Based on the effective variance of the strain field, we propose a quantitative method to identify the type of crack. The cracks can be identified as tensile crack, tensile and shear mixed crack, and shear crack, respectively, when the growth rate of the effective variance falls into the ranges between 0.72×10–2 and 1.89×10–2, between 2.34×10–2 and 3.59×10–2, and between 9.63×10–2 and 32.40×10–2, correspondingly.

Key words: rock mechanics, hole, flaw, digital image correlation method, effective variance, crack identification