Investigation of Spatter Characteristics in Selective Laser Melting Based on Maximum Entropy Threshold Segmentation Algorithm

The spatter morphology during selective laser melting (SLM) processing varies with process parameters, and it is difficult to achieve spatter extraction under all process parameters. The spatter extraction method based on traditional threshold segmentation only supports some process parameters and has no error analysis work , and the processing results cannot reflect the real spatter state. This paper proposes a robust image processing method to extract and process them based on the spatter images of the SIM process collected by high-speed cameras. The image processing method includes five steps, in which the threshold segmentation process depends on maximum entropy threshold segmentation algorithm. The results show that the spatter image processing method can accurately extract the spatter information under multiple process parameters. When the laser power is in the range from 100 W to 150 W , the change in the spatter area and number is determined by the molten state of the powder. And the reduction of spatter area and number is caused by spatter superposition when the laser power is in the range from 150 W to 200 W.