Research On Accurate Matching Of Joint Surface Of SLM Composite Manufacturing Based On Machine Vision

Turbine blades have parts with locally complex structures,which are difficult to process with traditional machining alone;it takes a long time to process simple structures with selective laser melting technology(SLM).For this reason,the composite manufacturing process that uses the combination of traditional machining and SLM technology can combine the advantages of the two processes to reduce the difficulty of processing such parts and shorten the processing cycle.When using composite manufacturing to process parts with partially complex structures such as turbine blades,the high-precision matching of the joint surface between the simple structure and the complex structure in the part is the guarantee of the accuracy of the final part.To this end,this paper takes the joint surface matching as the goal,and proposes a machine vision-based SLM compound manufacturing joint surface precise matching method.The main research contents of this paper are as follows:(1)A method for precise matching of the bonding surface of SLM composite manufacturing based on machine vision is proposed.Install the simple part of the traditional machined part on the substrate,collect the images of the two and obtain the position data of the joint surface based on the substrate plane coordinate system after image processing,and finally transmit the data to the SLM equipment to realize the joint surface matching.This method is man-made non-contact throughout the whole process,has high positioning accuracy,and will not cause damage to the equipment.(2)A set of substrate image acquisition system is designed.The internal size of the forming chamber of the SLM equipment is small,and the existing device cannot complete the substrate image collection work.For this reason,a set of image acquisition system with high stability and low cost was designed for the collection environment where the internal height of the forming chamber was 230 mm and the substrate diameter was 150 mm,and the various components in the system were selected reasonably to complete the physical construction.(3)A method of accurate secondary correction of the substrate image is adopted.After calibrating the internal parameters of the camera to correct the distortion of the substrate image,there is a phenomenon of incomplete correction at the edge of the image.Therefore,an image secondary accurate correction method is adopted.The pixel equivalent value at the optical axis of the camera is used as the unit.The high-precision standard checkerboard calibration board is used to establish a complete mapping matrix relationship for image correction,and the image is corrected twice and performed Error compensation.Experiments have proved that the data error in the image after secondary correction and error compensation is 91.43% within 17μm.(4)Obtain the positioning data of the joint surface matching.The metal particles in the forming chamber appear as salt and pepper noise in the image,so median filtering is used to reduce noise,and the filtered image is enhanced by histogram equalization,and finally the edges of the substrate and the target are extracted by the Canny algorithm.After determining the target’s center of gravity coordinates and the rotation angle of the feature point,the conversion relationship between the image pixel coordinate system and the substrate coordinate system is established,and the bonding surface positioning data is converted to the substrate coordinate system.(5)Designed a machine vision-based SLM composite manufacturing joint surface matching verification experiment.Establish circular and rectangular shape models with rotation angles,and perform the first sintering on the substrate.The sintering image is collected to extract the positioning data of the bonding surface of each shape model and input to the SLM equipment for the second sintering,and finally the error of the two sintering results is measured.In addition,the use of compound manufacturing to process the boss parts further proves the feasibility of the joint surface matching method in this paper.