Advanced Search
WANG Zhongren, XIA Hai, MA Xiaofeng, ZHANG Wen, CHEN Kepeng, ZHAO Rongli, SHI Tielin. Welding deviation measurement method based on molten pool image analysis[J]. TRANSACTIONS OF THE CHINA WELDING INSTITUTION, 2025, 46(8): 104-114. DOI: 10.12073/j.hjxb.20240528001
Citation: WANG Zhongren, XIA Hai, MA Xiaofeng, ZHANG Wen, CHEN Kepeng, ZHAO Rongli, SHI Tielin. Welding deviation measurement method based on molten pool image analysis[J]. TRANSACTIONS OF THE CHINA WELDING INSTITUTION, 2025, 46(8): 104-114. DOI: 10.12073/j.hjxb.20240528001

Welding deviation measurement method based on molten pool image analysis

  • During full-position gas arc welding at the consumable arc welding electrode of pipes, different azimuth angles of the pipe have different effects on the shape of the molten pool, and the trailing angle of the molten pool is closely related to the welding deviation. The welding deviation further affects the welding quality. Therefore, a welding deviation measurement method based on molten pool image analysis was designed. Firstly, after shooting the molten pool image, grayscale conversion and image enhancement were performed. The GR significance monitoring was used to segment the molten pool area, and then the Otsu threshold was used to extract the molten pool, detect the molten pool edge, and segment the contour. The sampling points of the molten pool trailing were extracted through straightness, and the linear fitting of the molten pool trailing contour was carried out using the gradient descent method. Finally, the trailing angle of the molten pool was calculated. To study the relationship between the trailing angle of the molten pool and the welding deviation, a linear regression method was used to design the experiment, and a linear regression mathematical model for predicting the trailing angle of the molten pool and the welding deviation was obtained. The experimental results show that the error between the predicted welding deviation and the actual welding deviation is within 0.2 mm, verifying the effectiveness of the model. Predicting the welding deviation in advance and correcting it in real time are of great significance for ensuring the welding quality.
  • loading

Catalog

    Turn off MathJax
    Article Contents

    /

    DownLoad:  Full-Size Img  PowerPoint
    Return
    Return