基于原位观测及机器学习的7050铝合金腐蚀行为表征  

Corrosion behavior characterization of 7050 aluminum alloy based on in-situ observation and machine learning

作　　者：牛瞳 张娜 熊希临 NIU Tong;ZHANG Na;XIONG Xilin(NCS Testing Technology Co.,Ltd.,Beijing 100081,China;Institute for Advanced Materials and Technology,University of Science and Technology Beijing,Beijing 100083,China)

机构地区：[1]钢研纳克检测技术股份有限公司,北京100081 [2]北京科技大学新材料技术研究院,北京100083

出　　处：《物理测试》2024年第1期1-7,共7页Physics Examination and Testing

基　　金：国家科技部重点研发计划(2021YFB3702204)。

摘　　要：原位观测了T7451及铸态7050铝合金在3.5%NaCl溶液中的腐蚀行为。利用扫描电镜(SEM)分析了腐蚀形貌特征及析出相成分,结合EBSD统计分析了晶粒取向与腐蚀行为之间的关系。并基于196条历史文献数据,运用Pearson相关性筛选及Backforward算法,对腐蚀行为特征进行重要性排序。发现7050铝合金腐蚀行为萌生于析出相周围,腐蚀深度及范围随时间增加,并伴有裂纹萌生。其中<112>、<114>、<324>3个晶粒取向上分布的析出相最多。通过动电位扫描表明,自腐蚀电流密度随析出相密度增加而增加。并且扫描电镜和机器学习重要性排序表明,Cu、Ti、Fe、Mg等元素均出现在析出相中,其对腐蚀行为的影响也最为显著。该工作能有效指导铝合金耐蚀性设计。The in-situ observation of corrosion behavior of T7451 and as-cast 7050 aluminum alloys in 3.5%NaCl solution was performed.The corrosion morphology feature and precipitated phase composition were analyzed by scanning electron microscopy(SEM),and the relationship between grain orientation and corrosion behavior was analyzed statistically by combining with EBSD.Based on 196 pieces of historical literature data,the importance ranking of corrosion behavior characteristics were conducted by Pearson correlation screening and Backforward algorithm.It was found that the corrosion behavior of 7050 aluminum alloy started around the precipitated phase,and the corrosion depth and range increased with time,which was accompanied with crack initiation.Among them,the precipitated phases were most distributed in the crystal faces of<112>,<114>and<324>.Potentiodynamic scan showed that the self-corrosion current density increased with the precipitated phase density.In addition,the importance ranking of SEM and machine learning showed that Cu,Ti,Fe and Mg all appeared in the precipitated phase,and their influence on the corrosion behavior was most significant.This study could effectively guide the corrosion resistance design of aluminum alloys.

关 键 词：7050铝合金 晶粒取向 腐蚀 机器学习 析出相 

分 类 号：TG146.21[一般工业技术—材料科学与工程]