Entwicklung eines Lebensdauervorhersagekonzepts im VHCF-Bereich auf der Basis kovariater mikrostruktureller Merkmalsgrößen

预期项目的目标是研究具有最大可能精度的疲劳寿命预测概念的统计模型和方法，以及在VHCF范围内的实验验证。已有的关于两种结构材料，镍基高温合金Nimonic 80A和奥氏体钢1.4301的变速疲劳性能的附加疲劳结果和微观结构分析作为数据库。所提出的数学研究工作一般可归结为考虑协变量信息的极值理论领域。协变量信息与控制失效的微观组织特征(如最大夹杂物尺寸、相邻颗粒的取向偏差、表面粗糙度等)有关。使用极值理论，应模拟试件疲劳寿命的条件分布，它依赖于外加应力幅值以及复协变量，并与所获得的数据进行拟合。因此，为建立适当的风险参数(如应力和失效循环次数)的良好接地证据作为众所周知的SN曲线的基础而创建了基础。统计模拟和实验调查之间的密切联系是该项目的核心部分。对于VHCF范围内的两种结构材料的实验结果应立即进入统计建模，并将所获得的结果用于实验设计。

The aim of the desired project is the study of statistical models and methods for a fatigue lifeprediction concept with the utmost possible accuracy as well as its experimental validation in theVHCF range. Already available and additional fatigue results and microstructural analysesregarding the VHCF behavior for two structural materials, the Ni-base superalloy Nimonic 80A undthe austenitic steel 1.4301, serve as database. The proposed mathematical research work cangenerally be attributed to the field of extreme value theory under consideration of covariateinformation. The covariate information relates to failure-governing microstructural features (e.g.maximum inclusion size, misorientation of neighboring grains, surface roughness etc.). Using theextreme value theory the conditional distribution of a specimen fatigue life depending on theapplied stress amplitude as well as complex covariates should be simulated and fitted to the dataobtained. Hence, a basis for the development of well-grounded evidences for appropriate riskparameters such as stress and number of cycles to failure as basis for the well-known SN-curve iscreated. The close link between the statistical simulation and experimental investigations is acentral part of the project. The experimental results for two structural materials in the VHCF rangeshould immediately enter into the statistical modeling, and the obtained findings should be used forthe design of experiments.