**Adrián Rodríguez-Panes, Juan Claver and Ana María Camacho \***

Department of Manufacturing Engineering, Universidad Nacional de Educación a Distancia (UNED), Madrid 28040, Spain; adrian.rodriguez@invi.uned.es (A.R.-P.); jclaver@ind.uned.es (J.C.)

**\*** Correspondence: amcamacho@ind.uned.es; Tel.: +34-913-988-660

Received: 2 July 2018; Accepted: 30 July 2018; Published: 1 August 2018

**Abstract:** This paper presents a comparative study of the tensile mechanical behaviour of pieces produced using the Fused Deposition Modelling (FDM) additive manufacturing technique with respect to the two types of thermoplastic material most widely used in this technique: polylactide (PLA) and acrylonitrile butadiene styrene (ABS). The aim of this study is to compare the effect of layer height, infill density, and layer orientation on the mechanical performance of PLA and ABS test specimens. The variables under study here are tensile yield stress, tensile strength, nominal strain at break, and modulus of elasticity. The results obtained with ABS show a lower variability than those obtained with PLA. In general, the infill percentage is the manufacturing parameter of greatest influence on the results, although the effect is more noticeable in PLA than in ABS. The test specimens manufactured using PLA perform more rigidly and they are found to have greater tensile strength than ABS. The bond between layers in PLA turns out to be extremely strong and is, therefore, highly suitable for use in additive technologies. The methodology proposed is a reference of interest in studies involving the determination of mechanical properties of polymer materials manufactured using these technologies.

**Keywords:** additive manufacturing; FDM; polylactide (PLA); acrylonitrile butadiene styrene (ABS); tensile behaviour; layer height; infill density; layer orientation

### **1. Introduction**

Additive manufacturing (AM) encompasses numerous technologies that allow for the construction of three-dimensional parts by superimposing layers of material. These techniques have undergone great development in recent years. The most widespread group of these is Material Extrusion manufacturing. These processes have the great advantage of low equipment and material costs, ease of use in laboratories and domestic environments, and the versatility to manufacture all kinds of shapes with a wide range of materials, mainly plastics, in a short time. Additive manufacturing enables the obtaining of extremely complex geometries in a single process, also with regard to interior cavities, thus providing great freedom during the design stage [1]. On the other hand, the development of new technologies and commercially available equipment for additive manufacturing using different materials, such as metals [2] and composites, brings the pieces obtained ever closer to the optimum values, not only with respect to performance in service, but also with respect to dimensional and geometrical precision [3]. Furthermore, this type of processes also provides advantages and new challenges [4] from the point of view of sustainability [5,6] and design optimisation [7,8]. Additive manufacturing is constantly developing and talk is already afoot regarding advanced concepts such as 4D printing, reflected by Momeni et al. in their 2017 review [9]. Its use is also being extended to

applications in medicine [10] and even within the area of university teaching, as shown in studies such as that carried out by García-Domínguez et al. [11].

The origin of the concept goes back to the 1980s with "Stereolithography" [12] and a whole host of new techniques have been developed since then, as explained by Gibson et al. in their book released in 2015 [13]. One of the categories or typologies acknowledged both in works undertaken by diverse authors [12] and in the recent ISO 17296-2:2015 standard on Additive Manufacturing [14] is Material Extrusion (Figure 1). This category encompasses processes in which the material, usually thermoplastic, is selectively applied using a nozzle to form each layer. The most important commercial process is known as Fused Deposition Modelling (FDM), which was patented by the founder of Stratasys over 20 years ago. Another similar process, but one that is not subject to this patent, is Fused Filament Fabrication (FFF).

**Figure 1.** (**a**) A sketch of the material extrusion process; (**b**) An extruder sketch in an FDM/FFF (Fused Deposition Modelling/ Fused Filament Fabrication) process.

When compared with conventional manufacturing processes, one of the main problems to be tackled is that the traditional test procedures cannot always be applied to additive manufacturing processes. As opposed to the traditional processing of polymers, layer-based production generates parts with anisotropic properties and residual stresses. For this reason, standardised methods are required that enable the linking of the properties of the material, the manufacturing parameters and the design of the piece, which represents a major challenge. Given the emerging nature of some of the technologies and their rapid development, the literature available in this area is not very extensive. However, a number of works have recently been published devoted to the mechanical characterisation of parts produced using FDM, such as that of Goh et al. [15], on reinforced thermoplastics; that of Domingo et al. [16], on using polycarbonate as the material; that of Mohamed et al. [17], where the effect of the process conditions on the temperature-dependent dynamic mechanical properties of processed PC-ABS (polycarbonate-acrylonitrile butadiene styrene) parts by FDM was analysed; that of Tymrak et al. [18], focusing on ABS and PLA (polylactide); and that of Chacón et al. [19], which concentrates on studying the influence of the manufacturing parameters on the mechanical properties of PLA parts, all of which show the interest of the scientific community to increase the knowledge about this subject.

Standard ISO17296-3 [20] covers the main surface, geometrical and mechanically requirements depending on the material (metal, plastic, or ceramic) and on the criticality ratings of the parts (highly engineered parts, functional parts that are not safety critical and prototype pieces). It also indicates the standards to be used to determine the principal quality characteristics and the corresponding

test standards, not only for the starting material, but also for the pieces produced. In this respect, the most appropriate test standard is indicated for determining the mechanical requirements of the plastic pieces produced using additive manufacturing, with ISO 527-2 [21] being specified for the tensile strength. It sets forth the tensile test conditions for rigid and semi-rigid extruded plastics, which is the closest example to the pieces that form the objective of this study. For its part, standard ISO/ASTM 52921-13 [22] specifies the nomenclature and terms associated with coordinate systems and testing methodologies for additive manufacturing. Additionally, on the other hand, standard ASTM D638-14 [23] establishes a test method for determining the tensile properties of reinforced and unreinforced plastics. The problem with all of these guidelines is that they refer to test procedures that still do not include specific considerations for AM. This lack of specific regulation with respect to additive manufacturing represents a serious problem as it makes it difficult to establish valid comparisons between machines, materials, and models that make it possible to predict the properties of the final pieces and establish design guidelines.

In this respect, Forster [24] collates the existing procedures for the testing of polymers and analyses their viability for additive manufacturing processes. The aforementioned standards ISO 527 [21] and ASTM D638-14 [23] can be consulted for the profiling of tensile properties. These standards are recognised as being valid for additive manufacturing processes, although amendments might be required with respect to the post-processing of the test specimens in order to meet the demands of the standard (surface finish or dimensional requirements) or their applicability might be limited as they do not meet the material isotropy requirements [25]. Although the methods for isotropic materials can be applied, this would lead to the results obtained being more uncertain and it will not be possible to equate the properties obtained for the material with those of the specific piece.

As claimed by Tymrak et al. [18], in order for FDM printed parts to be useful for engineering applications, the mechanical properties of parts produced by this technique must be known. Forster identifies a number of geometrical variables of the deposition of material with an influence on the mechanical properties of the piece, such as raster angle, the height and width of the layer, the space between extruded filaments, the combination of variables (space between extruded filaments, layer width and height or deposition velocity) that might increase the overlap between filaments, or the orientation of the piece during manufacture (which can affect the transfer of load between filaments and layer interfaces).

There are previous studies that have analysed the influence of some of these parameters for certain materials. Many of these focus on the influence of raster angle [26], a parameter of the process that significantly affects the anisotropy and strength of the pieces. According to the study undertaken by Rodríguez et al. [27], this parameter causes variations in Young's modulus between 11% and 37%. The majority of these works show that the parts are stronger when the lines are oriented in the load direction for tensile tests [28–30], mixed angles for flexion [28,31], and orthogonal to the yield load [32]. The higher the raster angle, the lower the tensile properties of the material [27,28], with these reaching a minimum of around 50◦ [33].

The mechanical strength of a piece produced using AM is always lower than that of the original material or that of injection moulded pieces [34]. However, 80% of an injection moulded piece's strength can be achieved by orienting the lines in the direction of the load [35]. Fernández-Vicente et al. [36] carried out a study that analyses the influence of the pattern and infill percentage, with the conclusion being that the influence of the different print patterns produces a variation of at least 5% in the maximum tensile strength, meaning that the performance is similar. On the other hand, the density of the infill is a decisive factor in tensile strength; the combination of a rectilinear pattern and a 100% infill provides the greatest tensile strength with a value of 36.4 MPa and a difference of less than 1% with respect to the filament (ABS).

Regarding the influence of the space between filaments, the presence of cavities and sharp corners increases the stress within the piece, which may cause failure [21]. In general, minimising the space between the filaments increases the contact area between them and lead to a stronger fusion interface. The material extrusion processes depend on the temperature gradients between contiguous filaments as these enable the thermoplastic polymers to form a solid fusion interface. Different studies set out to establish manufacturing guidelines related to the deposition speed and temperature and the temperature of the chamber which enable the production of stronger joints and, consequently, improved mechanical properties [32,37]. These type of studies enables the manufacturing parameters, the design of the parts and the final properties to be linked. Narrowing the width of the lines extruded reduces the residual stress in the filament and can increase the diffusion length. However, this would require more passes to create the piece, which increases the residual stress caused by the contraction of the polymer during cooling and, furthermore, the successive changes of nozzle speed have negative consequences in the diffusion [28].

With respect to manufacturing direction, Riddick et al. [29] combined orientations *xz*, *yz*, and *xy* with various raster angles (0◦, 0◦/90◦, and 90◦) and found that orientation *xz* had the highest modulus of elasticity, (*E* = 2.67 GPa) and the greatest tensile strength (15.26 MPa). Other authors have verified that the strength is increased by maximising the alignment of the layers in the load direction [31].

On the other hand, the manufacture of pieces using FDM/FFF processes is subject to numerous variables dependent upon thermophysical and/or chemical phenomena that are going to result in pieces with different characteristics depending on the method used along with the parameters of the process. One of these important phenomena is the inter-molecular diffusion between layers and/or dissimilar materials that influences the interfacial bonding strength, as explained in the work by Yin et al. [38] where inter-molecular diffusion theory based on heat transfer is developed and the influence of processing parameters on bonding strength has been investigated. A significant improvement in layer adhesion and a more isotropic part was obtained by Levenhagen and Dadmun [39] by developing a process in which bimodal blends of the same polymer with different molecular weight were used. Another interesting approach is the one presented by Ravi et al. [40], consisting of the development of a pre-deposition heating method to heat the region of an existing layer before the new layer is deposited; thus, the temperature at the inter-layer interface increases, improving the interpenetrating diffusion, leading to a better bond strength. Particularly for polymer-fibre composites but not limited to, other aspects including void formation, blockage due to filler inclusion and/or poor adhesion of fibres and matrix must be addressed in the close future, as stated by Parandoush and Lin in their work from 2017 [41]. The complexity of these processes due to the high number of parameters involved and their interdependencies requires multidisciplinary research [42]. Something that further complicates the analysis of the mechanical performance of these pieces is that they are normally fragile and fracture easily due to interlaminar failure caused by manufacturing defects that are difficult to control [15]. The nature of these phenomena is very complex, as they are also affected by the influence of related environmental and process conditions; for example, if the process is performed within a controlled atmosphere, the existence (or the lack thereof) of a heated bed and/or the heat transmission process affects the thermal gradients of the workpiece, particularly between layers.

Our study is focused on analysing more global trends for a wide range of parameters, including the most used values in the practice, independently of the equipment. The main interest of these parameters is that they can be accurately defined as they are strictly "manufacturing parameters" and they can be reproduced in all the FDM equipment independently of the technology used so that they can be reproducible for third parts in different application fields.

This paper sets out to establish a relationship between the manufacturing parameters and the mechanical properties of the piece by using the test procedures set forth in the existing standards on plastics for the polymer materials most used in FDM techniques: acrylonitrile styrene butadiene (ABS) and polylactide (PLA). These, together with nylon, polyethylene (PET, PETG), polycarbonate (PC), and thermoplastic elastomers (TPE) are commonly used for prototyping design and the creation of low-performance parts. This comparative study will enable us to know more about the mechanical performance of ABS and PLA pieces manufactured using FDM in terms of the main manufacturing parameters: the influence of layer height, the percentage of infill and the orientation of the object during manufacture.

#### **2. Materials and Methods**

#### *2.1. Materials and Equipment*

This paper analyses the mechanical performance of pieces manufactured using PLA and ABS, the most commonly used materials in material extrusion 3D printing technologies. PLA is a biodegradable polymer derived from lactic acid. The main advantage of this material is how easy it is to use in 3D printing and the good results it delivers. It requires a lower extrusion temperature than ABS, it does not suffer significant distortions during printing and it adheres well to the platform, which means it does not require a heated base. Neither does it give off a bad smell or toxic vapours during printing. It must not be used for parts that have to withstand high temperatures because PLA tends to warp at over 60 ◦C. ABS is a thermoplastic that is extremely resistant to impact, abrasion, and chemical elements. In 3D printing, it is the most used material after PLA. Its good mechanical properties, resistance to temperature, low price, moderate flexibility, long service life and range of melting temperatures make this material an excellent option for manufacturing all manner of parts using FDM technologies, above all, parts that have to withstand cyclical loads and temperature changes. However, it is not suitable for all applications as it presents problems of contraction and warping during printing, tends to peel away from the platform, and tends to give off toxic gases. Table 1 shows the data provided by the manufacturer for the filaments of the two materials used. As can be seen, PLA is more rigid and has a greater tensile strength while ABS is more ductile. However, the impact strength of ABS is far greater (320 J/m against 220 J/m (Notched Izod Impact)), one of the main properties that differentiates it from other plastics.

**Table 1.** The mechanical tensile properties of the filaments employed according to the manufacturer.


The additive manufacturing equipment to be used is the Prusa I3 Aluminium printer (Prague, Czech Republic) (Figure 2a), which employs FFF technology. The Cura software (Ultimaker, Geldermalsen, the Netherlands) is used to export the three-dimensional models of the samples to G-code. The main technical characteristics of the FDM printer are defined in Table 2. The tests are carried out using a HOYTOM HM-D 100 kN model universal testing machine (Leioa, Spain) (Figure 2b).

**Figure 2.** The equipment: (**a**) a Prusa i3 printer; (**b**) a HOYTOM HM-D 100kN Universal testing machine.


**Table 2.** The technical characteristics of the FDM equipment.

#### *2.2. Mechanical Properties of FDM Filaments*

Mechanical tests were carried out on filaments of the two materials in their raw state to compare the data with those obtained by the manufacturer. Two filaments of every material (ABS 1, ABS 2, PLA 1, and PLA 2) were tested by the authors (Figure 3) and the average values of the main mechanical parameters (tensile strength, nominal strain at break, modulus of elasticity) were compared with those provided by the manufacturer (Table 1) and presented in Table 3; in this case, the manufacturers did not specify how these properties were determined, but standardized specimens of thermoplastic materials for tensile testing are traditionally manufactured by injection moulding or machined. In this work, we want to check how accurate the results from those obtained by the manufacturer by traditional processes by applying tensile tests directly to filaments as they are the starting material for an FDM process. There are no specific standards for this kind of testing due to the novelty of the FDM technique, where pieces are obtained after depositing the filament during a layer by layer operation. The filaments are provided as solid wires of diameter 1.75 mm in coils and the tests were carried out directly by applying tensile forces to filaments of 165 mm in length.

**Figure 3.** The tensile tests of the filaments.

The mechanical properties derived from the tests are shown in Table 3.


**Table 3.** The comparison of mechanical properties (laboratory vs. manufacturer).

As it can be seen, the data obtained for the ABS and the PLA from tensile tests of the filaments are only qualitatively congruent with those provided by the manufacturer; the data that show a higher degree of coincidence are those associated with mechanical tensile strength, particularly results for ABS, that present a very similar value (0.4%); significant differences are found for the other two parameters recorded (nominal strain at break and modulus of elasticity), especially in the case of the material PLA, where differences of 112.5% and 44.6% have been obtained, respectively. This can be due to differences between the procedures followed, as explained before; however, the tensile strengths seem to have a better coincidence among results.
