Online Monitoring Method For SLS Based On Multi-Source Information

Additive Manufacturing has gained increasing attention from scientific communities due to its capability of manufacturing complex geometric parts,while the lack of quality control method restricted widespread application in industrial.In-situ online monitoring method becoming research hotpot in recent years is an effective way to enhance the capability of quality control and repeatability,but due to mostly used single sensor type the acquired information is localized and limited.The paper presents a multi-source information-based online monitoring method to Comprehensively take the advantages of infrared camera and depth camera and overcome the limitations of a single sensor in information acquisition for monitoring process for selective laser sintering(SLS).The main research content of this article is as follows:(1)According to the characters of the S320 selective laser sintering equipment and the features of infrared camera and depth camera installation method,a multi-source image registration method combining rough registration and fine registration is proposed.This method overcomes the shortcomings of the numerical iterative optimization method,such as long calculation time and slow convergence speed,and is simpler and easier to use in engineering applications.(2)On the basis of implementation and comparison,a multi-source image fusion method based on wavelet transform is determined.According to the GPU hardware structure and parallel computing theory,the fusion method is parallelized based on CUDA,greatly reducing computing time and meeting the real-time requirements of engineering applications.(3)A prototype of a multi source information online monitoring method was built,and the multi source information online monitoring system was used to analyze and check common surface defects of the powder bed during the selective laser sintering process.A method for detecting surface defects of the powder bed was determined based on the standard deviation of gradient amplitude.Based on temperature data,orthogonal experiments were designed to adjust the heating coefficient of the heating tube of the S320 equipment to ensure uniform temperature distribution.Using digital image processing methods and level set based active contour detection methods,the contour of the parts during the manufacturing process is extracted,laying a data foundation for subsequent active quality control based on the contour of the parts.The online monitoring method which uses infrared camera and depth camera comprehensively takes advantage of visible light image,depth image and infrared image,and based on image registration and image fusion methods,monitors the surface defects of the powder bed,temperature distribution uniformity and parts size during SLS process.The experimental results show that the standard deviation of the gradient amplitude of the powder bed surface defect image is more than twice that of the normal surface state,and the deviation of the histogram wave crest is more than 20%.The experimental results of the workpiece size monitoring show that the geometric contour extraction method of the workpiece can measure the size of the workpiece.Through the method,the temperature uniformity of the powder bed surface of S320 equipment was monitored,and the optimal temperature parameters were found through orthogonal experimental design.