Abstract

Metal material extrusion (MEX/M) is an accessible alternative for manufacturing metal components, although its industrial application is constrained, among other factors, by the dimensional and geometrical accuracy achieved. This Final Degree Project was carried out within the framework of the AEMMEX research project, led by the EPMAN group at the University of Oviedo, which aims to improve the accuracy of MEX/M-manufactured parts through geometrical compensation and data-driven optimisation.

The aim of this work was to analyse the influence of the main printing parameters on the dimensional and geometrical accuracy of green parts, i.e. parts before undergoing subsequent debinding and sintering treatments. To this end, a full factorial experimental design was developed involving four factors—layer height, extrusion multiplier, printing speed and extrusion temperature—with two levels per factor and three replicates, resulting in a total of 48 cubic specimens made of AISI 316L stainless steel. The specimens were manufactured using a Raise3D Pro3 printer and characterised using a coordinate measuring machine.

The results were statistically analysed through analysis of variance (ANOVA) and regression models. The results identified layer height and extrusion multiplier as the most influential factors affecting dimensional, form and orientation errors. The adjusted mean horizontal dimensional error was 0.617 mm, considerably higher than the vertical error of 0.096 mm. Over-extrusion was also identified as a common cause of reduced accuracy. Based on the analysis, simplified statistical models were developed and practical recommendations were formulated to improve part quality, prioritising lower layer heights and extrusion multiplier values, provided that process stability is maintained. This work provides an experimental and methodological basis for optimising manufacturing parameters and developing geometrical compensation strategies aimed at improving the accuracy of metal components manufactured using MEX/M.