PSI - Issue 56

Laszlo Racz et al. / Procedia Structural Integrity 56 (2024) 3–10

4

2

Author name / Structural Integrity Procedia 00 (2019) 000–000

1. Introduction Manufactured parts by fused deposition molding encounter several uncertainties regarding their mechanical proprieties, in fact due mainly to the formation of voids in the structure and inefficient bonding of layers. These aspects increase the damage probabilities in polymeric structures, the mechanical proprieties of the part being one of main disadvantages of this technology (Singh et al., 2020, Lalegani Dezaki et al. 2021). The most relevant factors that influence the mechanical characteristics of polymeric printed materials by FDM technology are related to the manufacturing parameters (infill rate, infill pattern, raster orientation, number of shells, layer height, speed while travelling, speed while extruding), the position of the element in relation to the printing platform, properties of the filament and other design characteristics of the 3D printer. In the literature there are several studies presenting the influence of printing parameters, such as infill ratio, infill pattern, layer thickness, layer height and other printer settings on the mechanical behavior of parts made by FDM technology (Bakır et al. 2021, Khan et al., 2021, Popescu et al. 2018, Dudescu and Racz 2017, Sajjad et al., 2023, Lalegani Dezaki et al. 2020). The accurate estimation of bonding between the filaments within a layer (termed "intra-layer bonding") and bonds formed between the filaments of the two succeeding layers (termed "inter-layer bonding") is essential to a reliable material model and strength evaluation of the parts made by FDM. The mechanical properties of such components are significantly influenced by the bonding quality between the filaments (Gurrala, et al. 2014, Sun et al. 2008). Theoretical calculations of the ultimate strength and E-modulus of printed structures typically need meso structure information and void density analyses. Additionally, the mechanical behavior of the FDM prototypes can be predicted using finite element analysis (Garg et al., 2017, Górski et al. 2015, Paul 2021). An FE model to evaluate inter-layer and intra-layer necking of the filaments during the diffusion of the raster layers throughout the printing process is presented by A. Garg et al. 2017. Another method proposed by Racz and Dudescu (2022) modelled the single filaments by a script using the G-code of the 3D printer. The connection between the filaments is established based on experimental analysis of the cross-sectional geometry of a printed tensile specimen and allows the estimation of the flattening effects of the filaments. The technique enables quantification of the E modulus of a printed tensile specimen with varying deposition densities (infill rate) as well as numerical estimation of the real cross-sectional area of a specimen and correction of the experimental stress-strain curves. The current study describes an experimental-numerical approach for simulating FDM 3D printed tensile specimens. To create the numerical model based on single filaments deposition, the approach relies on the 3D printer's original g-code. The bonding between the filaments is experimentally estimated by image analysis of the specimen’s cross section and allows the evaluation of the flattening effect that occurs during the deposition process. To evaluate the mechanical properties of the 3D printed material, uniaxial tensile tests are performed. When the material/airgap ratio is ignored the stress calculation can be easily done using the exterior dimensions of the tensile specimen. The presence of voids and the flattening effect of the filaments will generate a slightly different cross sectional area and will influence the engineering stress value computed after the tensile test. The presented methodology ensures a much accurate estimation of the real cross-section, tensile strength, and E-modulus of the investigated printed samples when the infill pattern inside it is changing. The current study analyses six different patterns which are mainly implemented in the printer’s software. 2. Materials and methods To determine the real cross-sectional area of the specimens with different infill rates, a numerical method has been employed (Racz, 2022, Garg et al., 2017). To create a realistic 3D model of the specimens, g-codes of the printed parts were used to build up the geometrical model. Based on the G-code, which was used for printing the specimens, a geometrical model was constructed in ANSA 17.1.2 (BetaCae, Greece). The virtual (CAD) tensile specimen (according to ISO 527-2-2012) was constructed in ANSA according to the process described above for six cases with the following infill patterns: grid 0°-90° and ±45°, triangular 60°, fast honeycomb, full honeycomb, and wiggle (Fig. 1). All models correspond to 100 % infill rate, so theoretical have full density.