PSI - Issue 28

James Allum et al. / Procedia Structural Integrity 28 (2020) 591–601 J.Allum et al. / Structural Integrity Procedia 00 (2019) 000–000

595

5

2.3. Mechanical and microscopic characterisation Tensile testing was undertaken using an Instron 5944 testing system with a 2 kN load cell. Testing was displacement-controlled with a rate of 0.5 mm min -1 with distance between the grips of approximately 20 mm (4.2 × 10 −4 s −1 strain rate). This distance was measured accurately using a digital caliper and recorded for calculation purposes before each specimen was tested. The cross-sectional area of each specimen was measured with optical microscopy to determine its load-bearing area. Seven microscopic measurements for each were collected (using ImageJ) and the mean value calculated. All microscopy was undertaken using a Zeiss Primotech optical microscope with a 5x magnification lens. 3. Results and discussion 3.1. Load-bearing capacity and load-bearing area Tensile tests showed that the F NG specimens (loaded in the direction of the extruded filaments) had the greatest load-bearing capacity with a mean peak load of 160 N. The manually grooved F G specimens all fractured in the region of manual grooving and had a mean load-bearing capacity of 140 N, while the Z specimens, loaded in the direction normal to the direction of extruded filaments, exhibited a mean load-bearing capacity of 123 N. magnitudes of load bearing area and load-bearing capacity were considered as these factors were directly influenced by the variation in geometry. The load-bearing capacity of all specimens is plotted in Fig. 4 as a function of mean load-bearing area. The latter was calculated from seven microscopic measurements of each specimen (as shown in Fig. 5). In a typical micrograph shown (Fig. 5) the delineated individual extruded filaments in F NG (a) and F G (b) are visible as they are oriented normal to the cross-sectional fracture surface shown. However, in the Z specimen (c), only the top view of a single interlayer interface is evident as the extruded filaments are aligned parallel to the surface shown, and thus hidden. The shaded area associated with the manually applied groove in F G is indicated on the micrograph (Fig. 5 (b)) along with the natural grooves in the Z specimen (Fig. 5 (c)). The microscopic measurements indicated that the manual groove in F G specimens resulted in a mean load-bearing area (2.10 mm 2 ), between that of the Z specimen (1.90 mm 2 ) and the non-grooved F NG specimens (2.39 mm 2 ). The load-bearing capacity and the microscopically measured mean load-bearing area were plotted (Fig. 4) for each specimen to find the relationship between load-bearing capacity and the cross-sectional area, over which the load was applied. Apparently (Fig. 4), as the area over which load was applied increased, the load-bearing capacity increased. This trend appears linear (Fig. 4), with all specimen types conforming to the same trend irrespective of their extruded filament orientations. Two of the specimens - F G and Z (circled in Fig. 4) - shared very close characteristics of load bearing areas and load-bearing capacities despite different extruded-filament orientations (one F and the other Z). This linear relationship (Fig. 4) is significant as it demonstrates that despite different extruded-filament orientations the load-bearing capacity depends on the geometry (load-bearing area) rather than orientation. This is evidence that the bond at the interface between layers in Z specimens is not the cause of reduction in load-bearing capacity, but rather it has the same mechanical properties as the bulk material. 3.2. Ultimate tensile strength The mean UTS for each specimen type was calculated and plotted (Fig. 6). All specimen types had very similar mean UTS values: 67.1 (F NG ), 66.8 (F G ) and 65.0 MPa (Z). The error bars (Fig. 6) indicate the range of values attained for each type, which was broader than the range of the mean values between all three specimen types. The mean UTS of bulk PLA as reported in the literature [15,23–29] was plotted for comparison. All specimens tested in this study had UTS values very similar to the mean reported for bulk material, irrespective of their extruded-filament orientation, presence (or not) of filament-scale geometric features or reliance on interfacial bonding. This further supports the finding of the previous section that the interlayer interface has bulk-material properties and, thus, the reduced strength of specimens loaded normal to the direction of extruded filaments was caused by filament-scale geometry so, a reduction in the mechanical properties of the interface bond when using common manufacturing parameters is likely caused by an ineffective measurement methodology or suboptimal manufacturing strategy.