Forming quality prediction of aluminum alloy self-piercing riveted joints based on deep learning
Article Text (Baidu Translation)
-
摘要: 为了解决模拟建模过程复杂、试验成本高、效率低的问题，提出了一种基于数据驱动的自冲铆接头成形质量预测方法。采用图像分割技术对试验获得的自冲铆接头截面图像进行预处理，生成了用于模型训练的接头成形截面类别区分图，详细标注了铆钉、上板、下板和背景的不同部分，为模型提供了丰富的视觉信息；基于试验图像获得模型的数据集，建立基于卷积神经网络的条件生成对抗模型体系结构的深度学习模型，将上下板材厚度与铆钉长度整合成一个包含3个标量的向量作为该网络模型的输入；在模型训练阶段，设计了15种不同的铆接工艺参数组合，分2个阶段对模型进行训练和预测，通过大量试验数据的训练和参数调整，模型的学习效果不断优化；在预测阶段，将模型的输出与试验获取的接头截面图像进行了对比分析，重点对标了钉脚张开度和残余底厚这2个关键几何参数。研究结果表明：训练后的深度学习模型可以准确预测不同基板厚度和铆钉长度组合的接头截面形状，钉脚张开度和残余底厚的平均预测精度分别为92.03%和92.48%；构建的自冲铆接头成形质量预测模型精度较高，提高了连接工艺的设计效率。Abstract: To address complex simulation modeling, high experimental costs, and low efficiency, a data-driven prediction method for the forming quality of self-piercing riveted (SPR) joints was proposed. Firstly, image segmentation technology was employed to preprocess the cross-sectional images of SPR joints obtained from experiments, generating categorized cross-sectional images for model training. These images meticulously labelled different parts of the rivet, upper plate, lower plate, and background, providing the model with abundant visual information. A dataset based on experimental images was established to build a deep learning model with a conditional generative adversarial network architecture based on convolutional neural networks. The thicknesses of the upper and lower plates and the length of the rivet were integrated into a vector containing three scalar values as the input for this network model. During the model training phase, 15 different combinations of riveting process parameters were designed, and the model was trained and predicted in two stages. Through extensive training with experimental data and parameter adjustments, the learning effect of the model was continuously optimized. In the prediction phase, the model's output was compared and analyzed with the experimentally obtained cross-sectional images of joints, with a focus on two key geometric parameters: rivet spread and residual bottom thickness. The results indicate that the trained deep learning model can accurately predict the cross-sectional shape of joints for various combinations of base plate thickness and rivet length, with average prediction accuracies of 92.03% and 92.48% for rivet spread and residual thickness, respectively. The developed model for predicting the forming quality of SPR joints demonstrates high accuracy, thereby enhancing the efficiency of the joining process design.
-
图 6 自冲铆接头成形截面图像预测模型收敛历史
Figure 6. Convergence history of the prediction model for the cross-sectional image of the self-piercing riveted joint
图 7 第1、2阶段真实图像和预测图像的对比
Figure 7. Comparison of real and predicted images in the first and second stages
图 8 第1阶段预测图像和试验图像的接头成形几何参数对比
Figure 8. Comparison of joint forming geometric parameters between predicted and experimental images in the first stage
图 9 第2阶段预测图像和试验图像的接头成形几何参数对比
Figure 9. Comparison of joint forming geometric parameters between predicted and experimental images in the second stage
表 1 十五组试验铆接工况参数
Table 1. Parameters of 15 sets of experimental riveting conditions
序号 铆钉长度/mm 上板厚度/mm 下板厚度/mm 1 4.0 1.1 1.3 2 1.1 1.5 3 1.1 1.7 4 1.3 1.3 5 1.3 1.5 6 4.5 1.1 1.3 7 1.1 1.5 8 1.1 1.7 9 1.3 1.3 10 1.3 1.5 11 5.0 1.1 1.3 12 1.1 1.5 13 1.1 1.7 14 1.3 1.3 15 1.3 1.5 
下载: 导出CSV
表 2 第1阶段的预测精度
Table 2. Prediction accuracy in the first stage
基板厚度/mm 预测精度/% 残余底厚 钉脚张开度 1.1+1.3 91.03 90.05 1.1+1.5 92.32 91.61 1.1+1.7 93.21 92.12 1.3+1.3 92.70 89.05 1.3+1.5 92.41 91.20 平均精度 92.33 90.81
下载: 导出CSV
表 3 第2阶段的预测精度
Table 3. Prediction accuracy in the second stage
铆钉长度/mm 预测精度/% 残余底厚 钉脚张开度 4.0 91.71 93.43 4.5 93.47 92.70 5.0 92.70 93.58 平均精度 92.63 93.24
下载: 导出CSV
-
[1] DU B, LI Q C, ZHENG C Q, et al. Application of lightweight structure in automobile bumper beam: A review[J]. Materials, 2023, 16(3): 967. doi: 10.3390/ma16030967 [2] 王琳, 胡光山. 汽车车身铝合金3层板自冲铆连接性能研究[J]. 塑性工程学报, 2024, 31(6): 95-102.WANG Lin, HU Guang-shan. Study on performance of self-piercing riveted joints of aluminium alloy three-layer sheet for automotive bodies[J]. Journal of Plasticity Engineering, 2024, 31(6): 95-102. [3] 李永兵, 马运五, 楼铭, 等. 轻量化薄壁结构点连接技术研究进展[J]. 机械工程学报, 2020, 56(6): 125-146.LI Yong-bing, MA Yun-wu, LOU Ming, et al. Advances in spot joining technologies of lightweight thin-walled structures[J]. Journal of Mechanical Engineering, 2020, 56(6): 125-146. [4] 欧阳潇, 秦登林, 赵升吨, 等. 无模无铆连接接头的塑性变形行为及工艺参数研究[J]. 塑性工程学报, 2024, 31(1): 41-49.OUYANG Xiao, QIN Deng-lin, ZHAO Sheng-dun, et al. Research on plastic deformation behaviors and process parameters of dieless clinching joints[J]. Journal of Plasticity Engineering, 2024, 31(1): 41-49. [5] 刘洋, 庄蔚敏, 何晓聪. 自冲铆接头成形及力学性能数值模拟关键技术研究进展[J]. 机械工程学报, 2022, 58(22): 168-185.LIU Yang, ZHUANG Wei-min, HE Xiao-cong. Research progress on key technology of numerical simulation of forming process and mechanical properties of self-piercing riveted joint[J]. Journal of Mechanical Engineering, 2022, 58(22): 168-185. [6] 游仁正, 窦炜, 张洪申, 等. 2024铝合金自冲铆成形工艺参数对接头微动滑移的影响[J]. 塑性工程学报, 2024, 31(11): 105-111.YOU Ren-zheng, DOU Wei, ZHANG Hong-shen, et al. Influence of process parameters of 2024 aluminum alloy self-piercing riveting on fretting slip of joint[J]. Journal of Plasticity Engineering, 2024, 31(11): 105-111. [7] ZHAO H, HAN L, LIU X P, et al. Comparisons of formation process and quality between two-layer and three-layer self-piercing riveted joints[J]. The International Journal of Advanced Manufacturing Technology, 2023, 127(9): 4745-4767. [8] TASSLER T, ISRAEL M, GOEDE M F, et al. Robust joining point design for semi-tubular self-piercing rivets[J]. The International Journal of Advanced Manufacturing Technology, 2018, 98(1): 431-440. [9] MA Y W, LOU M, LI Y B, et al. Effect of rivet and die on self-piercing rivetability of AA6061-T6 and mild steel CR4 of different gauges[J]. Journal of Materials Processing Technology, 2018, 251: 282-294. doi: 10.1016/j.jmatprotec.2017.08.020 [10] KAM D H, JEONG T E, KIM J. A quality study of a self-piercing riveted joint between vibration-damping aluminum alloy and dissimilar materials[J]. Applied Sciences, 2020, 10(17): 5947. doi: 10.3390/app10175947 [11] 张启森, 彭桂枝, 王晓莉. 基于正交试验的铝钢材料自冲铆接成形质量的多指标优化[J]. 锻压技术, 2020, 45(3): 87-91.ZHANG Qi-sen, PENG Gui-zhi, WANG Xiao-li. Multi-index optimization on forming quality of self-piercing riveting for aluminium-steel materials based on orthogonal experiment[J]. Forging & Stamping Technology, 2020, 45(3): 87-91. [12] ASATI B, SHAJAN N, SINGH ARORA K. Effect of process parameters on joint performance in self-piercing riveted dissimilar automotive steel joints[J]. Materials Today: Proceedings, 2022, 62: 721-726. doi: 10.1016/j.matpr.2022.03.658 [13] XIE Z Q, YAN W M, YU C, et al. Improved shear strength design of cold-formed steel connection with single self-piercing rivet[J]. Thin-walled Structures, 2018, 131: 708-717. doi: 10.1016/j.tws.2018.03.025 [14] LIU Y P, LI H, ZHAO H, et al. Effects of the die parameters on the self-piercing riveting process[J]. The International Journal of Advanced Manufacturing Technology, 2019, 105(7): 3353-3368. [15] 涂效铭, 金加庚, 李光耀, 等. 基于神经网络预测电磁自冲铆凹模匹配设计研究[J]. 塑性工程学报, 2025, 32(5): 43-51.TU Xiao-ming, JIN Jia-geng, LI Guang-yao, et al. Research on matching design of electromagnetic self-piercing riveting die based on neural network prediction[J]. Journal of Plasticity Engineering, 2025, 32(5): 43-51. [16] FERRÁNDIZ B, DAOUD M, KOHOUT N, et al. Prediction of cross-sectional features of SPR joints based on the punch force-displacement curve using machine learning[J]. The International Journal of Advanced Manufacturing Technology, 2023, 128(9): 4023-4034. [17] LIN J P, QI C W, WAN H L, et al. Prediction of cross-tension strength of self-piercing riveted joints using finite element simulation and XGBoost algorithm[J]. Chinese Journal of Mechanical Engineering, 2021, 34(1): 36. doi: 10.1186/s10033-021-00551-w [18] WU Q J, LIU Y, DAI Y L, et al. High-fidelity prediction of forming quality for self-piercing riveted joints in aluminum alloy based on machine learning[J]. Materials Today Communications, 2024, 41: 110319. doi: 10.1016/j.mtcomm.2024.110319 [19] FANG Y, HUANG L, ZHAN Z, et al. Effect analysis for the uncertain parameters on self-piercing riveting simulation model using machine learning model[J]. SAE Technical Papers, 2020, DOI: 10.4271/2020-01-0219. [20] CHEN S W, JIN D, HE H S, et al. Deep learning based online nondestructive defect detection for self-piercing riveted joints in automotive body manufacturing[J]. IEEE Transactions on Industrial Informatics, 2023, 19(8): 9134-9144. doi: 10.1109/TII.2022.3226246 [21] KARATHANASOPOULOS N, PANDYA K S, MOHR D. Self-piercing riveting process: Prediction of joint characteristics through finite element and neural network modeling[J]. Journal of Advanced Joining Processes, 2021, 3: 100040. doi: 10.1016/j.jajp.2020.100040 [22] ZHAO H, HAN L, LIU Y P, et al. Automatic and robust design for multiple self-piercing riveted joints using deep neural network[J]. The International Journal of Advanced Manufacturing Technology, 2022, 122(2): 947-975. doi: 10.1007/s00170-022-09893-8 [23] LI S M, YIN G Z, MA J Z, et al. Generation method for shaded relief based on conditional generative adversarial nets[J]. ISPRS International Journal of Geo-information, 2022, 11(7): 374. doi: 10.3390/ijgi11070374 [24] SHAFIQ M, GU Z Q. Deep residual learning for image recognition: A survey[J]. Applied Sciences, 2022, 12(18): 8972. doi: 10.3390/app12188972 [25] LIANG X L, XU J. Biased ReLU neural networks[J]. Neurocomputing, 2021, 423: 71-79. doi: 10.1016/j.neucom.2020.09.050 [26] WANG X Y, LIU J. Spatially regularized Leaky ReLU in dual space for CNN based image segmentation[J]. Inverse Problems and Imaging, 2024, 18(6): 1320-1342. doi: 10.3934/ipi.2024016 [27] RADFORD A, METZ L, CHINTALA S. Unsupervised representation learning with deep convolutional generative adversarial networks[J]. Computer Science, 2015, DOI: 10.48550/arXiv.1511.06434. [28] ISOLA P, ZHU J Y, ZHOU T H, et al. Image-to-image translation with conditional adversarial networks[C]//IEEE. 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). New York: IEEE, 2017: 5967-5976. [29] MAO X D, LI Q, XIE H R, et al. Least squares generative adversarial networks[C]//IEEE. 2017 IEEE International Conference on Computer Vision (ICCV). New York: IEEE, 2017: 2813-2821. [30] LEDIG C, THEIS L, HUSZÁR F, et al. Photo-realistic single image super-resolution using a generative adversarial network[C]//IEEE. 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). New York: IEEE, 2017: 105-114. [31] PATHAK D, KRÄHENBVHL P, DONAHUE J, et al. Context encoders: Feature learning by inpainting[C]//IEEE. 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). New York: IEEE, 2016: 2536-2544. [32] 刘小光, 朱伟庆, 周波, 等. 厚板T型接头焊接残余应力的形成机制与演变过程[J]. 交通运输工程学报, 2025, 25(4): 179-189. doi: 10.19818/j.cnki.1671-1637.2025.04.013LIU Xiao-guang, ZHU Wei-qing, ZHOU Bo, et al. Formation mechanism and evolution of welding residual stress in thick plate T-joint[J]. Journal of Traffic and Transportation Engineering, 2025, 25(4): 179-189. doi: 10.19818/j.cnki.1671-1637.2025.04.013 [33] BARAKAT A, BIANCHI P. Convergence and dynamical behavior of the ADAM algorithm for nonconvex stochastic optimization[J]. SIAM Journal on Optimization, 2021, 31(1): 244-274. doi: 10.1137/19M1263443 -
图(10) / 表(3)
计量
- 文章访问数: 119
- HTML全文浏览量: 54
- PDF下载量: 17
- 被引次数: 0
