Characteristic Of Weld Pool Detecting Of Plasma Deposition Direct Manufacturing

The technology of Plasma Powder Deposition Manufacturing (PPDM) integrates the plasma spray welding and finishing technology to manufacture the metal part or prototyping directly and rapidly. The control of welding quality is raised with be equal to the parameter control. The key is chrematistic detection of PDM. With chrematistic detection, welding quality is guaranteed, and techniques parameter is forecasted, thus the feedback of welding parameter is got. Detecting the weld line and its edge directly by visual sensor can achieve much useful information, which is very important in the research of weld line automatic tracking. To date, it is not fully used, because it is very difficult to detect clear and steady image owing to the noise interferences that exist in the nearby weld line caused by vast arc, spatter, etc. In according filter system is applied to remove the ultraviolet radiation and part visible light, and the remains is kept as light source, then Charge Coupled Device(CCD) camera is applied to obtain steady image of nearby weld line. The tasks such as image collection, storage, processing, characteristic detection and warp detection can be accomplished by the intercommunication between the hardware interfaces of image collecting equipment, servo driver and the computer image process unit.The arc can be shot directly by visual sensor after being processed by filter equipment. The advantage lies in that the image detected is two-dimension, which contains abundant and object information. In addition, the image processing arithmetic is practical. Thus, the image processing technologe obtains the beneficial factors of being applied in industry. By experiment analyze, the feasible extra filter system of PPDM is designed, and the approving result of filter optics was obtained.The image of weld line without large noise can be gained through image enhancement, image smoothness, and image filter and image edge enhancement of the original image of weld line. After the LOG filtering, a very clear weld line edge in front of weld pool was abstained. This provides convincing evidence for the further utilization of image processing in the weld line tracking and the hope to make further research. The max square difference auto threshold method and contour detection is used in this paper in the pretreatment of weld line image, through which the interference noise can be eliminated and distinct weld line image can be obtained. All of these build up a strong background for the weld line information detection. By the research of pixels grayscale of weld line image, image max square difference auto threshold and contour detection were used based on the characteristic of image grayscale straight square chart, and then, by making most use of weld line grayscale characteristic information. It can improve the dispose speed of weld line tracking procedure owing to its relative simple arithmetic. Through edge characteristic information identify arithmetic, the exact welding line edge information can be obtained, and the feedback of weld quality control can be feasible supplied.