Digital Twin of Fused Filament Fabrication Prints for Finite Element Analysis via G-Code Reverse Engineering

Abstract

As additive manufacturing by fused filament fabrication has gained popularity, computational analysis has become fundamental in predicting the mechanical behavior of 3D models. This paper proposes the development of a method for the finite element (FE) simulation of 3D-printed parts, implementing model design reverse engineering using G-code to obtain their digital twins (DTs). Samples were printed under the ASTM D638 standard with different nozzle diameters and layer heights, which allowed them to be mechanically characterized by tensile tests. The tensile tests determined that the diameter of the nozzles used (between 0.2 mm and 1.0 mm) influences the material’s tensile strength. The greater the diameter, the greater the stiffness, which translates into a change in the Young’s modulus, as well as greater tensile strength and thus a reduction of the deformation, for which a value of 2.66 $\pm$ $0.6\mathrm{\%}$ was obtained, i.e., the filament diameter did not influence this aspect. After carrying out the reverse engineering process of the samples to obtain DTs of the physical models, the printing G-code was used with the help of a Python script for their conversion to trajectories. These trajectories were introduced into Rhinoceros software with the Grasshopper add-on to obtain the reconstructed 3D models. The deposited filament profile used to reconstruct the DT was obtained by microscopy of the section of the physical samples. The predominant profile observed was that of a flattened oval. FE simulation was then carried out, obtaining a similarity of 90% between the simulated and mechanical tests, which validated the proposed method of predicting mechanical stresses in printed 3D elements.