Highly dense cellulose acetate specimens with superior mechanical properties produced by fused filament fabrication

Highlights

• Highly dense printed specimens.

• Mechanical performance similar to injection molded counterparts.

• New hot end modification.

Abstract

The extrusion based Fused Filament Fabrication (FFF) enables to build up components of thermoplastic materials by using a layer-by-layer strategy. Typically, the layered structure results in an internal porosity limiting the mechanical performance. In this work, using a commercial 3D-printer specimens of the polymer cellulose acetate (CA) were produced having porosities as low as 1–3 %. Quasi-static tensile tests demonstrate that these highly dense specimens reach similar or even higher strength values in comparison to their injection molded counterparts. The influence of the temperature management was studied by using different hot-ends, namely a standard one and a novel hot-end having a modified heat-block. Using the latter one, a homogeneous internal structure was achieved as observed with scanning electron microscopy (SEM). The very low porosities can be assigned to the elevated specimen temperature maintained during printing which is slightly above the materials glass transition temperature. Processing of CA at such temperatures leads to a slight loss of plasticizer as proved with NMR spectroscopy. This favorable change in material composition during printing additionally contributes to the mechanical strength of the test specimens. We conclude that the porosity is the key parameter (which has to be measured and minimized) to realize FFF components with good mechanical performance. Whereas we demonstrated FFF for the first time for CA as an industrially relevant thermoplastic material, our results can be easily applied to amorphous polymers in general.

Introduction

Additive Manufacturing (AM) is a fast growing technology field addressing not only prototyping and modelling [[1], [2], [3]], but also manufacturing products for end user applications [1,[4], [5], [6], [7], [8], [9]] as technical components, tools [7,9,10], spare parts [[10], [11], [12], [13]] and individualized consumer goods [[14], [15], [16]]. In particular, for individual construction, production of small numbers and components of complex geometries and functions AM can replace conventional (subtractive) manufacturing processes in e.g. medical engineering [9,17,18], automotive sector [6,7,19] and aerospace [4,5,20].

Fused Filament Fabrication (FFF) is a simple, versatile and low-cost AM-technique [[21], [22], [23]]. In FFF the raw material, usually being a filament of a thermoplastic, is fed into a hot-end. Within the hot-end the filament is molten followed by extrusion through a nozzle. The hot-end moves in the x-y-plane depositing the extruded strands on the (optionally heated) print bed. After one layer has been completed, the print bed is moved in z-direction followed by generating a new layer [24,25].

For end user applications it is crucial to manufacture components meeting all requirements for mechanical strength, surface quality and functionality. However, at present, the mechanical properties of fused filament fabricated components are significantly lower than their injection molded counterparts. The general notion about this lack of mechanical strength often refers to the so-called interlayer bonding. To be more precise, such an interlayer bonding results from the tight connection between adjacent strands, both in lateral and vertical direction. Such a connection may take place by a partial fusion of the strands in their contact zone, which is also called coalescence [[25], [26], [27], [28]]. The significance of an optimal interfacial adhesion between adjacent layers to reach high mechanical properties of printed components has been reported several times [[29], [30], [31]]. However, the degree of coalescence is difficult to quantify, and very often resulting properties of the printed components are taken as a qualitative measure. The degree of strand coalescence alters the geometry of the extruded polymer strands determining the amount of their contact surface as well as the proportion of voids (porosity) in the component. Therefore, the porosity is a suitable parameter to assess the interlayer bonding. Typically, porosities of specimens made with FFF range between 10 % and 60 % [[32], [33], [34], [35], [36], [37]]. However, many of the studies do not aim to achieve minimal porosities, as they are dealing with applications for medical engineering, e.g. surgical and diagnostic aids, prosthetics and tissue engineering where this is not of special concern [9]. Recently, Miller et al. reported values of less than 1 % porosity for FFF specimens made of polycarbonate urethane under optimized printing conditions such as a low printing speed. Although finding a mechanical performance for the FFF specimens similar to the injection molded samples, the authors did not analyze the physical background leading to these remarkable results [38]. For additively manufactured components using HDPE, Schirmeister et al. [27] achieved tensile strengths similar to those for injection molded counterparts. Levenhagen and Dadmun printed specimens of ABS and ABS blends with porosities of 2–7 %. The authors used a low molecular weight surface segregating additive, which significantly enhances the interlayer adhesion [31]. They obtained similar results with bimodal blends of PLA [29,30]. Blok et al. used carbon fibers to enhance the strength and stiffness of fused filament fabricated parts [28].

Numerous printing parameters have been investigated in order to understand and optimize the component strength. Commonly examined parameters [[39], [40], [41]] are process temperature [25,27,28,42], raster orientation [28,30,31,[43], [44], [45], [46], [47], [48]], component orientation [[49], [50], [51]] and layer thickness [[52], [53], [54], [55]]. Very recently, Schirmeister et al. printed HDPE and achieved remarkable mechanical strength values for nozzle temperatures above 200 °C, independently on the raster orientation [27]. In contrast, Blok et al. describes the raster orientation as being the main parameter to influence the strength of printed parts [28]. Very often, authors report significant differences in mechanical performance for varying the raster orientation [30,31]. The studies mentioned so far do not explicitly address porosity being affected by the printing parameters. Other authors investigated the porosity-strength relationship, but only for printed objects with high porosities. It is often suggested, that strand coalescence and low porosities are key parameters to achieve better mechanical performance [[32], [33], [34], [35], [36]]. However, a qualitative and quantitative understanding of the relation between porosity and mechanical strength is still lacking.

The most often studied materials in the field of FFF are PLA [29,30,56,57] and ABS [28,31,41,50]. Until now, Cellulose acetate (CA) is seldomly or not at all used in FFF. Only Pattinson et al. reported a 3D-printing technique processing CA from solution [58]. Others utilized CA based inks for the production of scaffolds [59,60]. As a thermoplastic material CA has good mechanical and optical characteristics. In particular, it is scratch-resistant and provides an excellent surface appearance with a shiny aspect. These properties enable the use of CA in many fields, e.g. tools, glasses, optical covers, etc. [61]. In comparison to other biomaterials, CA is readily available having a production capacity above 1 million tons per year [59]. It is an amorphous biopolymer, which is produced by an acetic acid process [62]. In this process, acetyl groups substitute the hydroxyl groups where the extent of this reaction is characterized by the degree of substitution (DS). In most cases, thermoplastic processing of CA is only possible by adding plasticizers with a total weight percentage of 15–40 % [[62], [63], [64], [65], [66], [67], [68]]. Typical plasticizers are phthalates (DEP, DMP), citrates (TBC, TEC), acetates (TA, EGDA), sebacates (DBS, DES), etc. [69]. For CA, depending on the plasticizer type and content, typical glass transition temperatures $T_g$ range from 70 °C to 120 °C [70].

Here, CA specimens were prepared with FFF using a novel hot-end modification [71]. Relevant printing parameters were varied such as raster orientation, component orientation and nozzle diameter. The porosities and the mechanical properties of the specimens were analyzed. Injection molded specimens served as a comparison. With this in-depth investigation the close relationship between porosity and mechanical performance is analyzed. Using nuclear magnetic resonance spectroscopy (NMR), differential scanning calorimetry (DSC) and thermogravimetric analysis (TGA) the pellets, the filament and the specimens produced were characterized to gain a comprehensive understanding of possible changes in material composition and material chemistry during processing.