PSI - Issue 49

Theodoros Marinopoulos et al. / Procedia Structural Integrity 49 (2023) 81–87 T. Marinopoulos et al./ Structural Integrity Procedia 00 (2023) 000 – 000

82 2

1. Introduction A prosthetic socket is a bespoke part that attaches a residual limb to the rest of the prosthesis. It needs to be fitted to each user, and its design depends on the case and the reason of amputation. Conventional prosthetic sockets are manufactured with the use of thermoplastics or reinforced polymeric resins. With the recent advances in additive manufacturing (AM), healthcare is where AM is finding more and more applications, for both upper- and lower-limb prostheses [1]. In the case of lower-limb prostheses, high loads must be sustained during their use, so it is crucial that their mechanical performance is sufficiently assessed before service [2,3]. Challenges arise due to the unique shape of each socket along with the lengthy manufacturing times, that make destructive tests expensive [4]. Because of that, very few reports can be found in the literature investigating the mechanical performance of prosthetic sockets, even less for those produced with AM. Additive manufacturing techniques can produce intricate one-off designs without the need for tooling and are therefore useful for producing anatomically shaped parts and products. AM methods can shorten the design to-manufacture cycle and reduce production costs, being adopted for several applications requiring anatomical shapes such as patient-specific implants, bone-repair fixation devices and prosthetic sockets [5,6]. However, AM parts often lack functional strength when compared to injection-molded or composite laminate equivalents, with most anatomical parts remaining at the prototyping stage due to weak interlaminar bonding and structural defects [2,7,8]. Some studies investigated the use of AM methods for prosthetic-socket manufacture [9 – 12], with material extrusion techniques being favoured thanks to their low cost and wide availability [13,14]. However, testing of such parts is not a standardised process; thus, few studies tested prosthetic sockets. Most commonly ISO 10328 standards [15] are followed as the closest indicator for tests since the socket should be able to bear the same weight as the rest of the components. The literature reports mixed results, with several AM sockets not meeting the minimum load requirements with a typical sudden interlaminar failure at loads much lower than those for thermoformed or composite-laminate equivalents [2,3,7,8,16]. Studies claimed to achieve acceptable strength for use as definitive sockets, employing custom machines for socket manufacturing; therefore, final products were expensive and not widely accessible, with compromises including design complexity, adaptability and dimensional accuracy [11]. AM paediatric sockets produced with standard 3D printers satisfying adjusted ISO 10328 loading requirements were also reported for above-knee sockets. It was found that increased contact area between layers in material-extrusion AM can lead in increased mechanical performance [17]. It was also observed that strategies based on generation of custom toolpaths reduced the number of grooves, improving the part strength [18] and the larger extruded-filament width increased the interlaminar load-bearing capacity by increasing the bonding contact area between adjacent filaments [19]. By bypassing conventional slicing software, custom toolpath strategies were also found to achieve more consistent mechanical properties in AM parts by reducing the number of stress raisers. In this study, it is hypothesised that bypassing the slicing software and automating the toolpath generation for smooth, concentric toolpaths with minimal stress raisers and maximum filament contact area will improve the mechanical performance of anatomically shaped parts such as prosthetic sockets. 1. Methodology The socket used in this study was an above-knee socket for the use by a 14-year-old boy with his weight and height in the 98 th percentile as used in previous studies [20,21]. Initially, slicing for printing was performed using Ultimaker’s commercial software CURA. The socket geometry was segmented with 0.25 mm layer height along its vertical axis. 100% infill was used, and the infill pattern “lines” was selected. A nozzle of 0.4 mm was used as the most commonly available and provided together with the printers. To study the effect of various slicing methods, alternative slicing techniques were investigated. This study employed open-source software FullControl [22]. Using an in-house developed Matlab code, the thickness of the socket walls was converted into concentric curves and, subsequently, concentric paths were created for the G-code. FullControl allowed the flexibility to change the width of the extrusion as a function of travel speed and feed rate along the created path. With wider extrusion, an increased contact area was achieved between layers. In this study, triple-pass socket walls were successfully created together with single-pass ones (Fig. 1). By adjusting the extrusion flow, for a constant melting temperature, it was possible to produce extra-wide paths up