Robotic AM system for plastic materials: tuning and on-line adjustment of process parameters

Industrial extruders mounted on robotic manipulators allow a fused material deposition rate that is 10 to 20 times higher than the average deposition rate of commercial FDM systems. Moreover, AM system based on robotic platforms could replace some of the application functions of FDM printers providing more flexibility, better motion software support and an industrial level of reliability. Eventually, the use of plastic pellets instead of wires results in a cost reduction and a higher freedom in material selection. Despite of these advantages, there are some drawbacks related to the manufacturing of big parts with high deposition rates, such as the irregular shape of deposited material in case of non-optimally tuned process parameters, which results in geometrical errors on the final part. Another critical issue is the material withdraw during the cooling phase, which could modify the deposited layer geometry. In the present study, an industrial screw-based extruder has been modified and mounted on an anthropomorphic robot, realizing a flexible platform for the additive manufacturing of big objects. This work will address the aforementioned limitations proposing a method to find optimal values for relevant process parameters and a method for online monitoring and control of process state-variables, thanks to the integration of sensors into the robotic system. In detail, in a first phase, a suitable experimental campaign has been developed according to Design of Experiments (DoE) in order to set the most important process parameters (extruder motor rotational speed, robot translation speed, layer height) ensuring a regular and constant deposited layer geometry. The relationship between the deposited track width and the aforementioned process parameters has been quantitatively studied by means of a statistical analysis of experimental results. In a second phase, a closed-loop control has been implemented to further improve the process parameter setting based on data measured during the deposition process, in this way compensating the material withdraw or other unexpected defects. The laser triangulation sensor, which has been mounted on the extrusion head, has been used to measure the actual height of each layer. Based on the acquired data, the robot path has been corrected by the closed-loop control to guarantee a proper layer overlapping and, therefore, a regular built-up geometry. A piece of furniture has been selected as representative case study of additive manufacturing of big parts and it has been manufactured to demonstrate the proposed procedure effectiveness.