可靠性(半导体)
蠕动
随机变量
结构工程
蒙特卡罗方法
概率逻辑
可靠性工程
计算机科学
振动疲劳
干扰(通信)
工程类
有限元法
材料科学
数学
统计
功率(物理)
物理
复合材料
人工智能
频道(广播)
量子力学
计算机网络
作者
Hang-Hang Gu,Run‐Zi Wang,Kun-Zhang,Ji Wang,Li Sun,Kai-Shang Li,Yu Liu,Xian‐Cheng Zhang,Shan‐Tung Tu
标识
DOI:10.1016/j.ijfatigue.2023.107768
摘要
Reliability assessment for structural system is of great significance in ensuring system safety, providing a scientific guidance for life management. In this work, a creep-fatigue reliability modeling method is developed for system-level perspective on the basis of cumulative damage-damage threshold interference. Probabilistic damage accumulations to given loading cycles are associated with the creep-fatigue models considering multi-source uncertainties and the damage threshold is regarded as a random variable. Particularly, a numerical procedure is proposed for practical application in engineering by combining the Monte Carlo simulation with Gaussian process regression-based surrogate model. Furthermore, a low-pressure turbine disk under creep-fatigue loading condition is taken as an instance to evaluate its reliability considering multiple weakness sites. Results show that the proposed method is better than the independent treatment because it is capable of avoiding the overly-conservative reliability estimation. In addition, the influence of damage summation rules and damage threshold on reliability is quantitatively analyzed in terms of the developed method. The effect of random parameters in structural system on computational efficiency of the proposed numerical procedure is investigated.
科研通智能强力驱动
Strongly Powered by AbleSci AI