Fused deposition modeling’s compressive behavior through a new finite element approach
摘要
Fused Deposition Modeling (FDM) is a subcategory of Additive Manufacturing (AM) technologies which has some essential manufacturing parameters, including raster angle, layer orientation, infill density, and nozzle diameter. These manufacturing parameters can affect the mechanical properties of a printed part. With this in mind, the parameters mentioned should be studied to find the optimum values. One of the most challenging tasks in the FDM technique is simulating the printed samples in the finite element analysis. Some researchers have recommended assuming isotropic assumption, and others proposed anisotropic one. However, anisotropic assumption needs more time and many basic experiments to obtain critical mechanical properties. Also, some researchers usually use solid elements to model FDM parts, and for modelling the true material properties, a reduction will be applied to cover the cavity percentage. On the other side, some researchers use anisotropic theories to model FDM parts which are complicated and costly. The current paper presents a new approach for modeling the FDM samples that need only the filament’s physical and mechanical properties for the first time. The novelty of the present work is to model a semi-discrete solid using beam elements and actual shapes of them in the finite element environment. This method not only reduce the finite element method’s run time, but also, increase the accuracy of the modeled parts’ responses. Here, some ABS disks printed by the FDM technique are modeled using the finite element method considering the filaments’ shape and intrinsic nature of them and compared with the experimental compressive test. Experimental data are derived from the previously published article about the effects of different path strategies on the compressive response of the FDM printed disks, and the numerical results are compared to those obtained experimentally. Finally, the approach used for finite element analysis can predict the failure load with a maximum error of 7% (Experimental failure load was 2955 N, while the finite element model predicted 3153 N). Also, a comprehensive discussion about the stress and strain distributions is reported. About the obtained discrepancies, the intrinsic nature of the FDM parts is discussed and some reasons are reported.