Experimental and numerical framework for characterization of high-strength concrete fatigue accounting for local dissipative mechanisms at subcritical load levels

近十年来，高强混凝土的疲劳响应已成为人们研究的热点。为了全面描述和解释疲劳损伤扩展的基本机理，并找出与正常强度混凝土的现象学差异，需要先进的数值模型和适当的实验方法。最近提出的疲劳损伤假说主要是由普通强度混凝土的累积应变测量驱动的，这一假说显示出很大的潜力来捕捉和解释胶凝材料的疲劳现象。因此，它被选为高强混凝土进一步发展和验证的基本理论概念。数值计算和实验研究将跨越三个层次的观测，包括局部破坏的非均匀损伤演化(级别1)、均匀损伤演化(级别2)和骨料间相互作用(级别3)。在级别1，骨料骨架的详细三维网格将用于研究格子离散模型(LDM)中骨料间滑移和微裂纹演化引起的损伤局部化。基于这些模拟，将设计一种新的实验方法，旨在分离在扭转和轴向联合载荷作用下的管状试件疲劳损伤演化的关键机制。开发的测试方法的一个显著特征是能够在测试区内诱导稳定的、宏观上均匀的疲劳损伤演化。在第二级，损伤过程的一致性将被用于模型的降维，这使得能够有效地模拟在高达数百万次循环的疲劳加载期间的三轴应力重新分布。微平面模型(MM)将被用来再现测试过程区域内均匀演化的损伤诱导的各向异性。在第三级，骨料间的相互作用将通过协调的数值模拟和测试来解决，以表征在水泥浆层内引起剪切或压缩的边界骨料结构。根据所描述的三个观察水平的覆盖范围以及局部机制的系统特征，将使用具有非均匀疲劳损伤演化的标准测试方法(压缩测试)来验证假设。这些结果将作为现实和有效地预测高强混凝土结构在一般荷载条件下的疲劳行为的基础，这些条件可以被纳入优先计划SPP2020内建立的实验-虚拟实验室。

Fatigue response of high strength concrete has become a subject of intensive research during the recent decade. To comprehensively describe and interpret the fundamental mechanisms governing the fatigue damage propagation and to identify the phenomenological differences with respect to normal strength concrete, advanced numerical models accompanied with appropriate experimental methods are needed. A hypothesis on fatigue damage primarily driven by a cumulative measure of strain postulated recently for normal strength concrete has shown a great potential to capture and explain the fatigue phenomenology of cementitious materials. Therefore, it has been chosen as the basic theoretical concept to be further developed and validated for high strength concrete within this research endeavor. The coordinated numerical and experimental research will span over three levels of observation including non-uniform damage evolution with localized failure (Level 1), uniform damage evolution (Level 2), and inter-aggregate interaction (Level 3).At Level 1, detailed 3D tessellations of the aggregate skeleton will be used to study the damage localization due to inter-aggregate sliding and microcrack evolution within an lattice discrete model (LDM). Based on these simulations, a new experimental method will be devised aiming to isolate the key mechanism of fatigue damage evolution on tubular specimens exposed to combined torsional and axial load. A salient feature of the developed test method will be the ability to induce a stable, macroscopically uniform fatigue damage evolution within a tested zone. At Level 2, the uniformity of the damage process will be exploited for dimensional reduction of the model, that allows for efficient simulations accounting for the three-axial stress redistribution during fatigue loading with up to millions of cycles. The microplane model (MM) will be used to reproduce the uniformly evolving, damage-induced anisotropy within the tested process zone. At Level 3, the inter-aggregate interaction will be addressed by coordinated numerical modeling and testing to characterize the borderline aggregate configurations inducing either shear or compression within the cement paste layer. Based on the described coverage of the three levels of observation with the systematic characterization of local mechanisms, the postulated hypothesis will be validated using standard testing methods with non-uniform fatigue damage evolution (compressive tests). The results will serve as a basis for a realistic and efficient prediction of fatigue behavior of high strength concrete structures exposed to general loading conditions that can be incorporated into the Experimental-Virtual-Lab established within the priority program SPP2020.