The Development Of Visual Monitoring System For Cutting Process Of Large Tube Setion Components

The hydrogenation reactor is a significant and critical equipment ofpetrochemical industry, whereas the hydrogenation tube section is a significantpart of the hydrogenation reactor. There are problems that existed in the turningprocess of tube section such as difficulties in chip breaking, chip winding causedtool wear and workpiece surface scratches etc., due to the thermal toughness andred hardness of the casting semifinished product of hydrogenation tube section.Thus, it is a great significanse of monitoring the turning process of tube sectionin consideration of increasing productivity hence the realization of the automatedproduction.The workpiece surface image based condition monitoring of the tool wear isthe significant part when monitoring the turning process of tube section.According to the working environment of the tube section turning process, thepresent research preliminary selected the front-end equipemnt for visialmonitoring and developed the visial monitoring system of the tube sectionturning process.The research analysed the workpiece surface formation as well as theinfluencing factors during the turning process, mainly studied the effect of toolwear on the workpiece surface texture, developed the relationship between thecharacteristic of the texture image and the tool wear, clarified the feasibility andrationality of workpiece surface texture based monitoring method of tool wearand completed the theoritical analysis and experimental research of featureextraction of the workpiece surface texture image as well as the discriminationmethod of tool wear feature.Commonly used image preprocessing method was studied combining with theworkpiece surface image observed from the experiment, and the pretretment ofthe surface image was carried out in the aspects of image geometric transformation, image correction, image segmentation, and image enhancement.An appropriate image preprocessing method of workpiece surface was developedthrough comparative analysis.The texture analysis of the workpiece surface image obtained from experimentwas carried out by applying the GLCM method, and characteristic parameters:second order matrix E, inertial matrix I, entropy H, dependency C etc. werechosen from the normalized GLCM as the characteristics of tool wear. From thecomparative analysis of these charateristic parameters, according to the inertialmatrix I with the orientation vertical direction between1, tool wear conditioncould be estimated and the monitoring of the tool wear condition could berealized, hence, a valid approch of tool wear monitoring during the machiningprocess was supplied.