Machine vision based statistical process control in fused deposition modeling

Abstract

Additive Manufacturing (AM) technology is widely used. There is an increasing demand for quality of AM product. Currently, there is no ideal method to monitor and ensure the quality of AM during the modeling process. In AM, the model is built layer by layer and the defects emerge gradually, so the defects might not be found immediately during the manufacturing process. With the accumulation of the inaccuracies, the quality of model would be decreased. To promote the quality level of AM, quality control technology is essential. In this paper, the quality of Fused Deposition Modeling (FDM) is studied by the method of combining the Machine Vision (MV) and Statistical Process Control (SPC). Firstly, the picture of each layer during the manufacturing process is caught with a camera. Secondly, the contour of these pictures are extracted with image processing methods. Finally, these data is analyzed with the methods of statistical process control and the control status is acquired. In this paper, three different models are designed to verify the feasibility of this method. It can be concluded that all defects in counter profile can be detected and the monitoring accuracy is 0.5mm.