-
不同于经典滚动接触疲劳形成的缺陷, 亚表面白蚀缺陷会引起轴承零件的早期失效, 严重缩短零件的寿命. 它位于金属亚表面且尺寸微小, 难以使用常规手段实现检测. 白蚀缺陷成因尚不明确, 不同演化阶段的缺陷样品制备耗时费力. 本文建立了白蚀缺陷演化模型, 基于k空间伪谱法开展了水浸超声检测过程数值实验. 对于含裂纹的白蚀缺陷演化后期, 可以忽略内部晶粒结构建立均匀层状模型, 使用经典声压反射系数幅度谱获取裂纹深度, 误差为1.5%. 对于不含裂纹的其他白蚀缺陷状态, 则存在内部声阻抗差异较小, 频谱特征不再明显等问题. 基于维诺图(Voronoi)建立轴承晶粒模型, 利用晶粒对超声的背散射效应来放大微观结构信号. 高频情况下, 基于深度卷积神经网络的训练准确率达92%, 验证准确率为97%. 即使在较低检测频率下, 背散射信号较弱, 仍能获得81%的准确率. 为白蚀缺陷的早期检测提供了有效方案.
Unlike classical defects formed by rolling contact fatigue, white etching defect (WED) including white etching area and white etching crack will cause surface to spall in the early stage and the service life to shorten seriously. Located in the subsurface of bearings, the tiny size WED is difficult to detect by conventional ultrasonic methods. The root cause of WED generation remains unclear. It is time consuming and expensive to prepare samples during the evolution of such defects. For characterizing the WED at early stage, five evolving states concerning the existing microscopic information are established in this paper. The immersion ultrasonic inspection process is simulated based on k-space pseudo spectrum method. For the later evolutionary stage with crack, the bearing can be simplified into a homogeneous three-layer model by ignoring the internal grain structure. The crack depth is obtained by using the ultrasonic reflection coefficient amplitude spectrum (URCAS), with an error of 1.5%. For other states without crack, the spectrum characteristic is no longer evident with slight acoustic impedance difference between layers. The polycrystalline structure on a microscale is thus realized based on Voronoi diagram, from which the grain induced backscattering can be used to amplify the microstructure variations at different stages. The backscattering signal is influenced by the grain size and detection frequency from the simulation. Since a direct comparison of backscattering information among evolutionary stages is difficult, the five different evolutionary stages of WED are recognized with the help of deep learning. The received waveform is transformed into a time-frequency map by short-time Fourier transform. Based on RESNET network structure, the results show that the train accuracy and validation accuracy reach 92% and 97% respectively. This study provides a sound way to characterize WED, which is conducive to early failure prediction and residual life evaluation. -
Keywords:
- white etching defect /
- ultrasonic backscattering /
- deep learning /
- ultrasonic reflection coefficient amplitude spectrum
[1] Manieri F, Stadler K, Morales-Espejel G E, Kadiric A 2019 Int. J. Fatigue 120 107Google Scholar
[2] Linzmayer M, Sous C, Gutiérrez Guzmán F, Jacobs G 2021 Wear 480–481 203925Google Scholar
[3] Leung J F W, Bedekar V, Voothaluru R, Neu R W 2019 Metall. Mater. Trans. A 50 4949Google Scholar
[4] Curd M E, Burnett T L, Fellowes J, Donoghue J, Yan P, Withers P J 2019 Acta Mater. 174 300Google Scholar
[5] Evans M H, Walker J C, Ma C, Wang L, Wood R J K 2013 Mater. Sci. Eng. A 570 127Google Scholar
[6] Lai J, Stadler K 2016 Wear 364–365 244Google Scholar
[7] López-Uruñuela F J, Fernández-Díaz B, Pagano F, López-Ortega A, Pinedo B, Bayón R, Aguirrebeitia J 2021 Int. J. Fatigue 145 106091Google Scholar
[8] Hu P, Turner J A, Tarawneh C, Wilson B, Fuller A J 2015 Proceeding of the Joint Rail Conference, San Jose, March 23–26, 2015 p5785
[9] Sreeraj K, Maheshwari H K, Rajagopal P, Ramkumar P 2021 Tribol. Int. 162 107134Google Scholar
[10] Ma Z, Zhang W, Gao J, Lin L, Krishnaswamy S 2016 43rd Annual Review of Progress in Quantitative Nondestructive Evaluation, Atlanta, July 17–22, 2016 p1016
[11] Ma Z, Zhao Y, Luo Z, Lin L 2014 Ultrasonics 54 1005Google Scholar
[12] Ma Z, Qi T, Lin L, Lei M 2022 Ultrasonics 119 106626Google Scholar
[13] 李珊, 李雄兵, 宋永锋, 陈超 2018 物理学报 67 107Google Scholar
Li S, Li X B, Song Y F, Chen C 2018 Acta Phys. Sin. 67 107Google Scholar
[14] Chen Y, Luo Z, Zhou Q, Zou L, Lin L 2015 Ultrasonics 59 31Google Scholar
[15] Norouzian M, Islam S, Turner J A 2020 Ultrasonics 102 106032Google Scholar
[16] 张永志, 辛全忠, 王永亮, 孔祥明, 刘昉, 杨再胜 2021 材料导报 35 24152Google Scholar
Zhang Y Z, Xin Q Z, Wang Y L, Kong X M, Liu F, Yang Z S 2021 Mater. Rep. 35 24152Google Scholar
[17] Liu H, Zhang Y 2019 Smart Mater. Struct. 29 015032Google Scholar
[18] Pyle R J, Bevan R L T, Hughes R R, Rachev R K, Ali A A S, Wilcox P D 2021 IEEE Trans. Ultrason. Ferroelectr. Freq. Control 68 1854Google Scholar
[19] Cai Y, Song Y, Ni P, Liu X, Li X 2021 Ultrasonics 117 106552Google Scholar
[20] Zhao Y, Lin L, Li X M, Lei M K 2010 NDT & E Int. 43 579Google Scholar
[21] He K, Zhang X, Ren S, Sun J 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Las Vegas, June 27–30, 2016 p770
[22] 郑战光, 汪兆亮, 冯强, 袁帅, 王佳祥 2016 广西大学学报(自然科学版) 41 460Google Scholar
Zhen Z G, Wang Z L, Feng Q, Yuan S, Wang J X 2016 J. Guangxi Uni. ( Nat Sci Ed) 41 460Google Scholar
[23] Weinzapfel N, Sadeghi F 2013 Tribol. Int. 59 210Google Scholar
[24] Bai X, Tie B, Schmitt J H, Aubry D 2018 Ultrasonics 87 182Google Scholar
[25] Yin A, Wang X, Glorieux C, Yang Q, Dong F, He F, Wang Y, Sermeus J, Van der Donck T, Shu X 2017 Ultrasonics 78 30Google Scholar
[26] Dryburgh P, Smith R J, Marrow P, Lainé S J, Sharples S D, Clark M, Li W 2020 Ultrasonics 108 106171Google Scholar
[27] Smith R L 1982 Ultrasonics 20 211Google Scholar
[28] 魏勤, 卫婷, 董师润, 张海林 2012 江苏科技大学学报(自然科学版) 26 27Google Scholar
Wei Q, Wei T, Dong S R, Zhang H L 2012 J. Jiangsu Univ. Sci. Technol. (Nat. Sci. Ed) 26 27Google Scholar
[29] Martin E, Jaros J, Treeby B E 2020 IEEE Trans. Ultrason. Ferroelectr. Freq. Control 67 81Google Scholar
[30] Gottlieb D, Tadmor E 1991 Math. Comput. 56 565Google Scholar
[31] Tillett J C, Daoud M I, Lacefield J C, Waag R C 2009 J. Acous. Soc. Am. 126 1231Google Scholar
[32] Loos J, Blass T, Franke J, Kruhoeffer W, Bergmann I 2016 J. Mech. Eng. Autom. 6 85Google Scholar
[33] 章桢彦, Toni B, 唐瑜, 韦剑飞 2020 中国冶金 30 2Google Scholar
Zhang Z Y, Toni B, Tang Y, Wei J F 2020 China Metallurgy 30 2Google Scholar
[34] 宋永锋, 李雄兵, 吴海平, 司家勇, 韩晓芹 2016 金属学报 52 378Google Scholar
Song Y F, Li X B, Wu H P, Si J Y, Han X Q 2016 Acta Metall. Sin. 52 378Google Scholar
[35] Bai X T 2021 M. S. Thesis (Xi’an: Xi’an University of Technology) (in Chinese) [白旭天 2021 硕士学位论文 (西安: 西安理工大学)]
[36] Behnke M, Guo S, Guo W G 2021 Proc. Manuf. 53 656Google Scholar
[37] Ha C, Tran V D, Ngo Van L, Than K 2019 Int. J. Approx. Reason. 112 85Google Scholar
[38] Mutasa S, Sun S, Ha R 2020 Clin. Imaging 65 96Google Scholar
[39] Song G, Qin D, Lyu Y, Hong G, Xu Y, Wu B, He C 2017 Int. J. Acoust. Vib. 22 511Google Scholar
-
表 1 模型使用的声学参数及晶粒尺寸
Table 1. Acoustic characteristics and crystalline grain size of different materials.
表 2 不同频率超声换能器的卷积神经网络计算结果
Table 2. Convolution neural network calculation results of different frequency ultrasonic transducer.
检测
频率/MHz回合 训练集
准确率/%验证集
准确率/%损失 检测
频率/MHz回合 训练集
准确率/%验证集
准确率/%损失 100 1 35.24 58.70 2.05 50 1 34.21 56.48 2.20 10 83.19 91.28 0.46 10 81.68 88.70 0.48 20 88.20 95.66 0.31 20 86.40 93.77 0.33 30 92.42 97.31 0.21 30 90.01 95.16 0.22 25 1 32.14 50.02 2.32 15 1 26.80 45.96 2.35 10 74.65 87.95 0.63 10 70.52 83.66 0.68 20 81.43 90.71 0.43 20 79.92 88.50 0.45 30 85.99 93.40 0.22 30 81.03 91.02 0.22 -
[1] Manieri F, Stadler K, Morales-Espejel G E, Kadiric A 2019 Int. J. Fatigue 120 107Google Scholar
[2] Linzmayer M, Sous C, Gutiérrez Guzmán F, Jacobs G 2021 Wear 480–481 203925Google Scholar
[3] Leung J F W, Bedekar V, Voothaluru R, Neu R W 2019 Metall. Mater. Trans. A 50 4949Google Scholar
[4] Curd M E, Burnett T L, Fellowes J, Donoghue J, Yan P, Withers P J 2019 Acta Mater. 174 300Google Scholar
[5] Evans M H, Walker J C, Ma C, Wang L, Wood R J K 2013 Mater. Sci. Eng. A 570 127Google Scholar
[6] Lai J, Stadler K 2016 Wear 364–365 244Google Scholar
[7] López-Uruñuela F J, Fernández-Díaz B, Pagano F, López-Ortega A, Pinedo B, Bayón R, Aguirrebeitia J 2021 Int. J. Fatigue 145 106091Google Scholar
[8] Hu P, Turner J A, Tarawneh C, Wilson B, Fuller A J 2015 Proceeding of the Joint Rail Conference, San Jose, March 23–26, 2015 p5785
[9] Sreeraj K, Maheshwari H K, Rajagopal P, Ramkumar P 2021 Tribol. Int. 162 107134Google Scholar
[10] Ma Z, Zhang W, Gao J, Lin L, Krishnaswamy S 2016 43rd Annual Review of Progress in Quantitative Nondestructive Evaluation, Atlanta, July 17–22, 2016 p1016
[11] Ma Z, Zhao Y, Luo Z, Lin L 2014 Ultrasonics 54 1005Google Scholar
[12] Ma Z, Qi T, Lin L, Lei M 2022 Ultrasonics 119 106626Google Scholar
[13] 李珊, 李雄兵, 宋永锋, 陈超 2018 物理学报 67 107Google Scholar
Li S, Li X B, Song Y F, Chen C 2018 Acta Phys. Sin. 67 107Google Scholar
[14] Chen Y, Luo Z, Zhou Q, Zou L, Lin L 2015 Ultrasonics 59 31Google Scholar
[15] Norouzian M, Islam S, Turner J A 2020 Ultrasonics 102 106032Google Scholar
[16] 张永志, 辛全忠, 王永亮, 孔祥明, 刘昉, 杨再胜 2021 材料导报 35 24152Google Scholar
Zhang Y Z, Xin Q Z, Wang Y L, Kong X M, Liu F, Yang Z S 2021 Mater. Rep. 35 24152Google Scholar
[17] Liu H, Zhang Y 2019 Smart Mater. Struct. 29 015032Google Scholar
[18] Pyle R J, Bevan R L T, Hughes R R, Rachev R K, Ali A A S, Wilcox P D 2021 IEEE Trans. Ultrason. Ferroelectr. Freq. Control 68 1854Google Scholar
[19] Cai Y, Song Y, Ni P, Liu X, Li X 2021 Ultrasonics 117 106552Google Scholar
[20] Zhao Y, Lin L, Li X M, Lei M K 2010 NDT & E Int. 43 579Google Scholar
[21] He K, Zhang X, Ren S, Sun J 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Las Vegas, June 27–30, 2016 p770
[22] 郑战光, 汪兆亮, 冯强, 袁帅, 王佳祥 2016 广西大学学报(自然科学版) 41 460Google Scholar
Zhen Z G, Wang Z L, Feng Q, Yuan S, Wang J X 2016 J. Guangxi Uni. ( Nat Sci Ed) 41 460Google Scholar
[23] Weinzapfel N, Sadeghi F 2013 Tribol. Int. 59 210Google Scholar
[24] Bai X, Tie B, Schmitt J H, Aubry D 2018 Ultrasonics 87 182Google Scholar
[25] Yin A, Wang X, Glorieux C, Yang Q, Dong F, He F, Wang Y, Sermeus J, Van der Donck T, Shu X 2017 Ultrasonics 78 30Google Scholar
[26] Dryburgh P, Smith R J, Marrow P, Lainé S J, Sharples S D, Clark M, Li W 2020 Ultrasonics 108 106171Google Scholar
[27] Smith R L 1982 Ultrasonics 20 211Google Scholar
[28] 魏勤, 卫婷, 董师润, 张海林 2012 江苏科技大学学报(自然科学版) 26 27Google Scholar
Wei Q, Wei T, Dong S R, Zhang H L 2012 J. Jiangsu Univ. Sci. Technol. (Nat. Sci. Ed) 26 27Google Scholar
[29] Martin E, Jaros J, Treeby B E 2020 IEEE Trans. Ultrason. Ferroelectr. Freq. Control 67 81Google Scholar
[30] Gottlieb D, Tadmor E 1991 Math. Comput. 56 565Google Scholar
[31] Tillett J C, Daoud M I, Lacefield J C, Waag R C 2009 J. Acous. Soc. Am. 126 1231Google Scholar
[32] Loos J, Blass T, Franke J, Kruhoeffer W, Bergmann I 2016 J. Mech. Eng. Autom. 6 85Google Scholar
[33] 章桢彦, Toni B, 唐瑜, 韦剑飞 2020 中国冶金 30 2Google Scholar
Zhang Z Y, Toni B, Tang Y, Wei J F 2020 China Metallurgy 30 2Google Scholar
[34] 宋永锋, 李雄兵, 吴海平, 司家勇, 韩晓芹 2016 金属学报 52 378Google Scholar
Song Y F, Li X B, Wu H P, Si J Y, Han X Q 2016 Acta Metall. Sin. 52 378Google Scholar
[35] Bai X T 2021 M. S. Thesis (Xi’an: Xi’an University of Technology) (in Chinese) [白旭天 2021 硕士学位论文 (西安: 西安理工大学)]
[36] Behnke M, Guo S, Guo W G 2021 Proc. Manuf. 53 656Google Scholar
[37] Ha C, Tran V D, Ngo Van L, Than K 2019 Int. J. Approx. Reason. 112 85Google Scholar
[38] Mutasa S, Sun S, Ha R 2020 Clin. Imaging 65 96Google Scholar
[39] Song G, Qin D, Lyu Y, Hong G, Xu Y, Wu B, He C 2017 Int. J. Acoust. Vib. 22 511Google Scholar
计量
- 文章访问数: 3852
- PDF下载量: 85
- 被引次数: 0