ABSTRACT

This paper reports pitting corrosion loss data of AA5005-H34 aluminium alloy immersed in natural seawater for up to 2 years. It is shown that the data for mass loss, maximum pit depth and the average value of 15 deepest pits as a function of exposure time are not closely consistent with the classical power-law function. Instead, the data show a greater affinity to the early part of a bi-modal trend. The uncertainty of the pit depth data was analysed using extreme value theory. The results are that scatter in the data sets increases with exposure time. This is considered to be the result of the pit depth data population being non-homogeneous, characterised by a mixture of deep pits with differing pitting morphologies. The results of this study suggest that longer-term data and homogeneous data population are likely to be more reliable for future corrosion loss prediction purposes.

摘要

本文报告了AA5005-H34铝合金在天然海水中浸泡长达2年的点蚀腐蚀损失数据。结果表明，质量损失，最大凹坑深度和15个最深凹坑的平均值作为暴露时间的函数的数据与经典幂律函数并不十分一致。相反，数据显示出对双峰趋势早期的更大亲和力。使用极值理论分析了坑深数据的不确定性。结果是，数据集中的散布随曝光时间的增加而增加。这被认为是凹坑深度数据总体不均匀的结果，其特征是深浅凹坑具有不同的凹坑形态。