Correlation Between The Change Behavior Of The Melt Pool And The Occurrence Of Defects In SLM

Selective Laser Melting(SLM)is one of the most promising additive manufacturing technologies with the advantages of high precision,high density and good mechanical properties.However,the forming quality is difficult to control due to the influence of powder material and processing parameters.Various defects during processing have become one of the biggest bottlenecks limiting the development of SLM technology.Online monitoring of SLM forming process is the key means to solve this problem,and is also an important prerequisite to achieve part quality traceability and real-time feedback control.In this paper,based on the independently built high-speed imaging monitoring system and layer-wise imaging monitoring system,the melt pool change behavior,powder coiling and denudation,and defect detection of the forming layer are investigated.The main works are as follows:(1)For the melt pool and spatter image,the melt pool area is extracted by using the image processing methods of threshold segmentation,morphological operation and edge detection;the characteristics of melt pool area,perimeter and roundness are counted,and the change pattern of melt pool characteristics of single-track and multi-track parts without and with powder is studied;the change behavior of melt pool is analyzed,and the different formation mechanisms of bulge defects on single-track and multi-track parts are investigated by combining the microscopic images of melt path morphology.(2)For the powder coiling and denudation effect caused by melt pool change,the movement pattern of powder particles during the single track forming process is studied;the effects of the process parameters on the powder coiling speed and the width of denudation are statistically analyzed;the intrinsic connection between the powder coiling phenomenon and the occurrence of poor melting and crater defects is investigated through online analysis of the multi-track part forming process and offline analysis of its surface morphology.(3)Based on the above melt pool scale study,the defect monitoring at the powder bed scale is further investigated.For the three types of forming layer defects:geometric deformation,debris and local bulge caused by melt pool change and powder coiling,a feature fusion-based defect detection method is proposed,which extracts edge features,texture features and geometric features of forming layer image using histogram of oriented gradients,gray level co-occurrence matrix and Hu invariant moment respectively;a new feature vector is obtained by weighted serial fusion of the three features,and the Support Vector Machine(SVM)is combined to achieve high accuracy and low time consuming defect detection of forming layer.