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

592

2

1. Introduction In recent years MEAM (Material Extrusion Additive Manufacturing) has seen rapid development and interest from major manufacturing companies globally. Recent advances broadened the capabilities and accessibility of the technology. MEAM provides a number of benefits compared to traditional manufacturing methods, including the capability to generate bespoke parts with complex geometries at a relatively rapid rate. This is particularly relevant to high-value industries such as medicine, where the investigation and development of patient-specific implants such as hip joints [1] or cranial plates [2] are of increasing interest. Mechanically, anisotropy is broadly recognised as the biggest limitation in MEAM. It was widely attributed to the influence of the layer-by-layer manufacturing process, which results in numerous interfaces between layers, responsible for poor mechanical performance in the transverse direction (normal to direction of extruded filaments) [3–13]. Typically, studies showed that the reduced performance was associated with deficiencies in bond formation due to limitations in manufacturing parameters and thus attempted to find ways of improving bonding by parametric optimisation. Broadly, thermal modifications (nozzle temperature, printing speed) or geometrical modifications (layer height, raster pattern) were considered, but this led to numerous contradictions. For example, studies showed that reducing the layer height improved mechanical performance [4,5,9,14–17]; on the contrary, improved mechanical performance as layer height was increased was also demonstrated [5,13,14,16,18]. Likewise, studies demonstrated improved mechanical performance both with increased [9,18] and reduced [13,14,16] print speeds. Variation in mechanical behavior may relate to the complexity and variation of specimen designs utilised. As yet, there are no standardised testing methods for MEAM: so, most studies adapted existing polymer-testing standards for characterisation, such as ASTM D638 [19]. Obviously, such standards were developed for moulded specimens and were not intended for characterisation of MEAM parts, with their inherent microscale variation (filament-scale geometric features) and complex thermal history owing to deposition sequencing. Differences in the interpretation of standards resulted in disparity in methodological approaches, giving rise to discrepancies in microscale geometry and thermal history of tested specimens, making it very complicated (and in some cases impossible) to directly compare findings across different studies. This led us to develop a new microscale testing specimen, manufactured with reproducible identical printing conditions. Recent studies by the authors [20–22] demonstrated that reducing geometrical and thermal complexity enabled more accurate and precise understanding of the fundamental mechanical properties of MEAM specimens. This was made possible by development of our novel test specimens for improved, fundamental analysis. The dogbone specimen, comprising only single extruded filaments through its thickness and generated with a symmetrical toolpath, reduces much of the geometrical and thermal complexity found in other studies. A microscopy methodology enabled precise quantification of the load-bearing regions in the specimen’s cross-sections (rather than by using nominal or digital-caliper measurements), providing new understanding of mechanical properties and demonstrating that the interlayer interface has the strength of the bulk material. Utilising this specially developed test specimen, this study further investigates the claim of bulk-strength interfaces, by manually applying grooves to a longitudinal (F direction) tensile-test dogbone specimens to mimic the filament scale geometric grooves, naturally present in the transverse (Z direction) specimens due to the interlayer interface geometry. This study directly examines the cause of anisotropy by characterising the mechanical performance of the interlayer interface (Z direction) specimen against both the unchanged F non-grooved specimen (F NG ), and the manually modified F grooved (F G ) specimen. The manually grooved specimen has a filament-scale geometry similar to that of the Z specimen but crucially composed of the longitudinally deposited filaments. A comparison of the fracture characteristics and tensile properties of the specimens should indicate whether the presence of the grooves is responsible for the interfacial weakness at the interlayer interface or whether there is a fundamental material weakness caused by bonding, between the layers.