基于反馈神经网络的机身镀层抗热冲击性能研究  

Thermal shock resistance of fuselage coating based on feedback neural network

作　　者：刘利[1,2] 钟林[1] LIU Li;ZHONG Lin(School of Artificial Intelligence&Big Data,Luzhou Vocational and Technical College,Luzhou 646000,China;Southwest Jiaotong University,Chengdu 610031,China)

机构地区：[1]泸州职业技术学院人工智能与大数据学院,四川泸州646000 [2]西南交通大学,四川成都610031

出　　处：《兵器材料科学与工程》2023年第1期106-111,共6页Ordnance Material Science and Engineering

摘　　要：用磁控溅射离子镀设备在机身材料为高速钢M2的试件上制备Cr、Ti、N含量不同的4种镀层。用HADL2101DS8动静态电阻应变仪测试4种镀层在500℃热冲击下的水平方向、厚度方向的应力,输入反馈神经网络,计算不同镀层的破损率,评价其抗热冲击性能,确定最佳的镀层制备方案。用扫描电子显微镜观察镀层受热冲击前、后形貌。结果表明:当Cr、Ti、N的质量分数分别为56.89%、0.05%、43.06%时(方案1),镀层形貌无裂缝或脱落、氧化物白色颗粒最少,镀层破损率小于另外3种方案;方案1镀层抗热冲击性能最优。Four coatings with different contents of Cr,Ti and N were prepared on high-speed steel M2 using magnetron sputtering ion plating equipment.The HAD-L2101DS8 dynamic and static resistance strain gauge was used to measure stression the horizontal and thickness direction of four kinds of coatings under thermal shock at 500℃.The results of stresses were input into the feedback neural network to calculate the damage rate of different coatings,evaluate their thermal shock resistance,and determine the best preparation scheme of coating.The morphologies of the coatings before and after thermal shock were observed with a scanning electron microscope.The results show that when the ratios of Cr,Ti and N are 56.89%,0.05%and 43.06%respectively(scheme 1),the coating morphology shows no cracking or peeling,and the white oxide particles are the least,and its percentage of damage is lower than the those of other three schemes.Scheme 1 shows the best thermal shock resistance.

关 键 词：CrTiN镀层 机身镀层 抗热冲击性能 反馈神经网络 

分 类 号：TQ171[化学工程—玻璃工业]