Development of a real-time layer height estimation technique during directed energy deposition based on infrared thermographic imaging

Abstract

The challenging problem of layer height monitoring during a directed energy deposition (DED) process is crucial for the assurance and control of product dimensions and quality in additive manufacturing. In this study, a real-time layer height estimate method based on infrared thermographic imaging is developed specifically for the DED process. The suggested system estimates the layer height by entering process parameters and measured melt pool properties into an artificial neural network (ANN). The following are examples uniqueness and advantages of the developed technique: (1) An ANN is developed and optimized for the real-time estimation of layer height during the DED process. (2) Both numerical and experimental analyses are carried out to examine the correlation between layer height and melt pool properties. (3) The proposed system can estimate the layer height even under the complex movement of the printing nozzle due to the coaxial design of the installed infrared camera, making it suitable for additive manufacturing of complex geometries. (4) A single infrared imaging system can estimate the width, length, depth, and layer height of a melt pool at the same time when combined with the authors' method for estimating melt pool depth. The overall root mean squared error (RMSE) and absolute percentage errors in height estimate for an average layer height of 200 μm were 25.44 μm and 12.6%, respectively.