中文摘要：

采用试件养护前期拔出预埋插片形成预制裂隙的方法制作多裂隙脆性材料试件,并在伺服控制单轴加载系统上对其进行加载试验。基于单压下脆性材料局部破损的应变软化机理建立裂隙体数值模型,对比试验与数值计算结果分析有序多裂隙脆性材料破坏机制及其影响因素。结果表明：除裂隙倾角及其几何排布外,裂隙在试件中的相对分布位置也影响裂隙体的破坏特征,且影响程度与裂隙面上有效剪切力大小有关,有效剪切力越大,则影响越显著。试验与数值计算结果显示：多裂隙试件中存在一组优势破坏面,与和裂隙走向相一致的试件斜对角线重合,裂隙分布在优势破坏面上或附近时,其尖端发育微裂纹的机率大于远离优势破坏面位置裂隙。结合有序多裂隙试件破坏特征及数值模型单元屈服状态,提出2种近置裂隙尖端裂纹发育简化模型,并结合本试验结果验证了简化模型的可行性。

英文摘要：

The loading tests were done on the brittle specimens with ordered multi-fissures which were made by pulling out the metallic shims at the early stage in the curing process under a uniaxial loading system with servo control.The numerical models of fissure bodies were established based on the strain-softening mechanism of brittle material with local damage characteristics under uniaxial compression.The failure mechanism and its influencing factors of brittle material with ordered multi-fissures were analyzed by comparing experiment results with numerical calculations.The results show that the failure characteristics of specimens are also influenced by the relative position of fissures besides the fissure inclination angle and its geometrical arrangement,and the influence degree has relationship with the effective shear stress on fissure surface.The larger the effective shear stress is,the more significant the influence is.The experimental results and numerical calculations show that a dominant failure surface exists in the specimen with ordered multi-fissures,which is coincident with the diagonal that has the same strike with fissures.The probability of micro-cracks appearing at the tips of fissures,which distribute in or are close to the dominant failure surface,is larger than the probability of the ones which are far from the dominant failure surface.Combined with the failure characteristics of specimens with ordered multi-fissures and yield situations of grid cells in numerical models,two kinds of simplified models are proposed to describe the spatial morphologies of closely-spaced fissures.Then,combined with this experimental result,the feasibility of simplified models is verified.