A novel melt pool mapping technique towards the online monitoring of Directed Energy Deposition operations

Abstract

Although Directed Energy Deposition (DED) has more than a two and a half decade history, the difficulty of consistently controlling this process is still limiting its adoption. Achieving robust printing of arbitrary shapes with high-fidelity (macro-scale) geometry and uniform pre-determined micro structure throughout relies on careful coordination of many factors (some of which still rely on the knowledge of the system user). Key to this endeavor is controlling the activity of the melt pool and managing its influence on surrounding material (including due to reheating cycles). Even with emerging in-process melt pool control methods, translating ideal steady state melt pool models into practical use while creating arbitrary shapes on-the-fly can be difficult to visualize and validate. Despite recent advances in the field of melt pool control, intuitive visualization of melt pool signatures and confirmation of how they affect the quality of the deposition is yet to be found. A process control indicating possible flaws or defective regions in the material component, as well as certifying the uniformity of characteristics during deposition would be of great help to improve the process itself and further diffuse its application. This study presents the development of a novel and innovative method for acquiring, processing and visually showing DED process signatures to assess progress toward fully automated melt pool control. This methodology relies on incrementally building a map of actual process conditions by merging multiple data streams collected during the deposition process for each layer and then plotting the recorded signatures along the 2D toolpath position. This concept has been realized with a new software application called HeatMAP. This application uses data supplied by the AMBIT™ Melt Pool Measurement ‘MPM’ system wherein images from the melt pool are acquired by a CMOS camera and processed. A second data stream is simultaneously collected directly from the CNC including the actual feed speed and the nozzle positions by using the DTConnect software (an application developed in this study based on FANUC FOCAS library). A visualization of the merged data sets in the HeatMAP software then imparts to the user an intuitive way of assessing the process quality using visualization as an indicator of metallurgical quality. To demonstrate the utility of this approach, two sets of experiments were undertaken: one as a benchmark, where process signatures were captured by the MPM system in monitoring only mode and another with active closed-loop laser control. When the melt pool measurement system is used in closed loop mode, the laser power is varied in order to maintain a target melt pool area within a specific range. At present, image data was acquired and processed throughout two deposition strategies, which were used to prove the mapping concept, its capabilities and methodology. The methods behind HeatMAP and DTConnect presented in this study, demonstrate the potential for these applications to accelerate the development of more stable deposition parameters, and process planning and control in the future. Using the proposed control, improvements to material quality are expected in the DED process.