Elsevier

Composites Part B: Engineering

Volume 113, 15 March 2017, Pages 371-380
Composites Part B: Engineering

Influence of meso-structure and chemical composition on FDM 3D-printed parts

https://doi.org/10.1016/j.compositesb.2017.01.019Get rights and content

Abstract

We studied the influence of extruded filament dimensions and chemical composition on mechanical behavior of Fused Deposition Modeling (FDM) objects made of Acrylonitrile Butadiene Styrene (ABS) polymer. All aspects are investigated through experimental campaigns: we tested three different filament dimensions on the same material and two different chemical compositions with the same filament dimensions.

We verified that FDM ABS specimens show anisotropic mechanical properties since they vary with filament extrusion direction. Accordingly, Classical Lamination Theory (CLT) and Tsai-Hill yielding criterion were found to be well capable of predicting in-plane stiffness and strength of FDM specimens. We assessed that, varying chemical composition and filament dimensions, it is possible to tune fiber properties and fiber-to-fiber bonding and, consequently, the overall mechanical properties at macro-scale, in particular the yielding strength and the strain at failure. The experimentally obtained data are useful to calibrate mechanical models to be used with computational tools as finite element analyses.

Introduction

Additive Manufacturing (AM) processes form three-dimensional (3D) objects from virtual models, obtained from a Computer Aided Design (CAD) software, digital scanning systems or medical imaging systems, such as computer tomography or magnetic resonance imaging. In recent years, AM processes have begun to progress from rapid prototyping techniques towards rapid manufacturing methods, where the objective is now to produce finished components [5]. AM technologies are ready to be used for industrial production and, due to a growing competition between service furnishers, they are becoming economically feasible for a great number of end-user applications [15]. In the last decade the maturity of these processes was largely increased thanks to the research on materials, the development of new equipments and the better understanding of the processes [27].

From an industrial point of view, technologies capable of producing robust parts with high strength and long-term stability are the most relevant, because they allow the direct production of end-user parts. The use of AM solutions for direct production would be possible only if mechanical properties are well known and taken into account in the design stage, depending on the process parameters. However, one of the most important open issues in rapid manufacturing is the prediction of 3D-printed parts behavior under real working conditions.

In this study we focus on mechanical properties of Fused Deposition Modeling (FDM®) 3D-printed objects. Such process belongs to the material extrusion subfamilies of Solid Freeform Fabrication (SFF) technologies. In FDM processes a thermoplastic filament is heated and extruded through a robotically controlled head: the material is deployed layer by layer on a printing surface in a temperature controlled environment. Each printed layer is made of filaments known as fibers (also called beads or roads) deposited in a plane parallel to the printing surface. The printing head movements, the extrusion system and all the other printing parameters are controlled by an electronic board, relying on a set of instructions (G-Code) listed in a file. The G-Code is produced by a dedicated software commonly called slicer or slicing software, that takes into account the virtual geometry, the characteristics of the printing material and the specific features of the 3D-printer.

The FDM 3D-printed object is composed by two main parts, the internal raster (infill) and its outer shell, made by perimeters and solid top/bottom layers; the direction of the deposited material is known as fiber orientation angle. The bonding between neighboring fibers occurs by a thermally-driven diffusion process during solidification of the semi-molten extruded fiber [22], [6].

The inner structure at a sub-millimeter scale, i.e. the meso-structure, is determined by the filament path deposition and process parameters. Among these, the most important are: fiber thickness and width, fiber-to-fiber overlap, fiber orientation and extrusion temperature. Fig. 1 shows a schematic representation of the ideal meso-structure resulting from the FDM process.

Research efforts in FDM technology have been directed towards the evaluation of the mechanical properties of the resulting part as a function of process parameters. Ahn et al. [1] showed that both fiber-to-fiber overlap and fiber orientation had a significant effect on the resulting tensile strength, while compressive strength was not affected by these parameters. Sood et al. [25], [24] investigated the functional relationship between fiber dimensions, overlap and orientation and specimen strength using response surface methodology; results show that such parameters influence fiber bonding capabilities and distortion of the printed part and, consequently, the compressive strength of test samples. Moreover, Lee at al. [17], [18] studied the influence of fiber thickness, orientation and fiber-to-fiber overlap on the strength of FDM 3D-printed Acrylonitrile Butadiene Styrene (ABS) samples; the authors determined that the compressive strength for an axial FDM specimen is greater than for a transverse one. Rodríguez et al. [21], [22] compared elastic modulus and tensile strength of FDM printed samples with the same properties of the ABS filament feedstock; the authors concluded that parts with fibers aligned with the axis of the tension force have the greatest tensile strength. Es-Said et al. [8] investigated the influence of fiber orientation and polymer molecules alignment along the extrusion direction.

Concerning the mechanical modeling of 3D-printed parts, Classical Lamination Theory (CLT) and Tsai-Hill yield criterion have already been considered in few studies. Kulkarni and Dutta [16] applied CLT to describe the elastic moduli of FDM printed laminates. Bertoldi et al. [7] and Rodríguez et al. [23] assumed orthotropic material symmetry and obtained elastic moduli and strength values for different fiber orientations. Li et al. [19] studied the fabrication process and the mechanical properties of FDM specimens, using CLT to determine the elastic constants as a function of raster angle; experimental data were in good agreement with the results of the laminate modeling.

Literature evidence proves that CLT and Tsai-Hill yielding criterion are valuable instruments to describe the mechanical properties of FDM 3D-printed components. Accordingly, in the present work, we use both instruments to investigate how the mechanical properties of ABS 3D-printed FDM parts are influenced i) by the chemical composition of the filament and ii) by fiber cross-sectional dimensions, maintaining the thickness/width ratio constant. The influence of filament dimension and chemical composition on mechanical behavior is studied varying fiber orientation with respect to the loading direction. All aspects are investigated through experimental campaigns: meso-structure influence, i.e. fiber thickness and width is tested on the same material, while chemical composition impact is tested using the same meso-structure.

Section snippets

Elastic behavior modeling

In this study, FDM 3D-printed objects are considered as composite laminates consisting of orthotropic laminas; each lamina corresponds to a layer made of ABS parallel filaments.

We consider two right-handed coordinate systems, namely {O,x,y,z} and {O',1,2,3} as in Fig. 2: axis 1 is aligned to the filament extrusion direction and represents the longitudinal direction of the lamina, while axis 2 is perpendicular to fiber extrusion direction and represents its transverse direction. Axes x and y

Preliminary material analysis

Two different types of ABS, provided by Versalis S.p.A. [11] and indicated with letters A and B, are investigated: type A is the base ABS, while type B is a chemically additivated ABS. The 3D printer used for the present study is a 3NTR A4v3 [12]. The machine is equipped with three extruders, which can be heated up to 410 °C, through a ceramic heating component; a nozzle of 0,4 mm of diameter is used. The bed temperature can reach 120 °C, while the heated chamber can reach 75 °C. Preliminary

Results

Tensile tests on specimens at 0 are able to capture the mechanical response of a 3D-printed part that mainly depends on the mechanical behavior of the fiber i.e. on intra-fiber properties. Conversely, tensile tests at 90 are suitable for retrieving the mechanical characteristics of 3D-printed parts which primarily depend on the bonding process, i.e. on inter-fibers properties. During tensile tests we observed that specimens, especially longitudinal ones, exhibited whitening regions, i.e.

Discussion

Elastic modulus validation data (20 and 70 specimen) show a good consistency with theoretical estimation for configuration A1 and B1, while for configurations A2 and A3 some overestimation of CLT with respect to experimental data has been observed (see Fig. 5).

Validation data for Tsai-Hill criterion are also well consistent with theoretical estimation, with some slight underestimation for 70 specimens of configurations A1 and B1 (see Fig. 6): the reason probably lies in the air gap

Conclusions

In this paper we investigated how fiber orientation, filament dimensions and chemical composition affect the mechanical properties of ABS 3D-printed components. In particular, we tested and compared i) three different meso-structures on the same material and ii) two different types of ABS using the same meso-structure.

As already highlighted by previous works, we verified how a 3D-printed material shows anisotropic mechanical properties. Accordingly, CLT and Tsai-Hill yielding theory were found

Acknowledgements

The presented activity is inserted in the framework of 3D@UniPV (http://www.unipv.it/3d), one of the strategic research area of the University of Pavia.

References (27)

  • M. Baldinger et al.

    Price benchmark of laser sintering service providers

  • C. Bellehumeur et al.

    Modeling of bond formation between polymer filaments in the fused deposition modeling process

    J Manuf Process

    (2004)
  • Bertoldi, M., Yardimci, M. A., Pistor, C. M., Guceri, S. I., Sala, G., 1998. Mechanical characterization of parts...
  • Cited by (145)

    View all citing articles on Scopus

    The results of this work have been presented at the 2016 International Workshop on Multiscale Innovative Materials and Structures (MIMS16), Cetara (Salerno), Italy, 28–30 October 2016.

    View full text