Abstract The tunnel-type anchorage (TTA) is recently being used in large-span suspension bridge, however, the reliability is not yet investigated and the pullout load carrying behavior is also not fully understood. Hence, in this paper, the pullout behavior of TTA during loading was investigated using a finite element program. Subsequently, extensive parametric studies were carried out to assess the factor of safety for TTA. Based on the study results, a predictive model for the safety factor of TTA was developed via an artificial neural network approach, and then the predictive model was implemented into a spreadsheet containing an automated search algorithm to calculate the reliability probability of TTA. The calculation results were evaluated using Monte Carlo simulation. The effects of the design parameters of plug body and its variability on the reliability of the TTA were discussed. Besides, a series of implications for design are also proposed. Some useful conclusions regarding the stress distribution and evolution of TTA during loading as well as the reliability are presented.

中文翻译：

---

隧道式锚固稳定性的可靠性评估

摘要 隧道式锚固（TTA）近年来在大跨度悬索桥中得到应用，但其可靠性尚未得到研究，拉拔承载行为也未完全了解。因此，在本文中，使用有限元程序研究了加载过程中 TTA 的拉拔行为。随后，进行了广泛的参数研究以评估 TTA 的安全系数。基于研究结果，通过人工神经网络方法建立了TTA安全系数的预测模型，然后将该预测模型实施到包含自动搜索算法的电子表格中，以计算TTA的可靠性概率。使用蒙特卡罗模拟评估计算结果。讨论了塞体设计参数及其可变性对TTA可靠性的影响。此外，还提出了一系列对设计的影响。介绍了一些关于加载过程中 TTA 的应力分布和演变以及可靠性的有用结论。