摘要
为评估9310合金钢镀层的抗腐蚀能力,提出基于人工神经网络的9310合金钢镀层抗腐蚀能力分析方法。以双真空熔炼的9310钢为原材料,结合两种金属涂层粉末,通过超音速火焰喷涂工艺制备WC-14Co涂层(涂层1)和WC-10Co4Cr涂层(涂层2)这两种样品。对两种样品进行酸性盐雾干湿循环腐蚀试验,观察腐蚀失重及微观形貌变化。将涂层工艺参数输入到BP-ANN(反向传播人工神经网络)模型中,BP-ANN模型通过学习样本数据,挖掘涂层工艺参数与抗腐蚀能力之间隐藏的内在规律和特征,实现精确预测。结果显示,在腐蚀28 d时,两种样品表面微观形态差异不显著,仅生成少量的腐蚀产物;而当腐蚀时间延长至80 d时,两个样品表面均出现明显的腐蚀坑。BP-ANN模型预测效果良好,决定系数值分别高达0.982和0.986。经腐蚀失重预测确定了最佳工艺参数取值,其中涂层样品1最小失重为1.66 mg,涂层样品2最小失重为1.13 mg。这些验证了BP-ANN预测模型在预测9310合金钢镀层抗腐蚀能力方面的有效性。
To comprehensively evaluate the corrosion resistance of 9310 alloy steel coating,a method for analyzing the corrosion resistance of 9310 alloy steel coating based on artificial neural network was proposed.WC-14Co coating(Coating 1)and WC-10Co4Cr coating(Coating 2)were prepared by supersonic flame spraying process using 9310 steel melted in double vacuum as raw material and two types of metal coating powders.Acid salt spray dry wet cyclic corrosion tests were conducted on two samples to observe corrosion weight loss and microstructural changes.Input the coating process parameters into the BP-ANN(Backpropagation Artificial Neural Network)model,enabling it to learn sample data and uncover the underlying patterns and features between coating process parameters and corrosion resistance,achieving accurate prediction.The analysis results show that there is no significant difference in the surface microstructure between the two samples after 28 days of corrosion,and only a small amount of corrosion products were generated.When the corrosion time was extended to 80 days,obvious corrosion pits appeared on the surfaces of both samples.In addition,the BP-ANN model has good prediction performance,with determination coefficient values as high as 0.982 and 0.986,respectively.Through corrosion weight loss prediction,the optimal process parameter values were determined,with the minimum weight loss of coating sample 1 being 1.66 mg and coating sample 2 being 1.13 mg.These results validate the effectiveness of the BP-ANN prediction model in predicting the corrosion resistance of 9310 alloy steel coatings.
作者
王波
元尼东珠
Wang Bo;Yuanni Dongzhu(School of Information Engineering,Jiaxing Nanhu University,Jiaxing 314001,China;School of Computer,Qinghai Nationalities University,Xining 810007,China)
出处
《电镀与精饰》
北大核心
2025年第12期63-70,共8页
Plating & Finishing
基金
青海省科技厅的基金(2019-ZJ-7066)。
关键词
人工神经网络
9310合金钢
镀层
抗腐蚀能力
腐蚀失重预测
表面微观形态
artificial neural network
9310 alloy steel
coating
corrosion resistance
corrosion weight loss prediction
surface microstructure
