This study investigates the effect of filament moisture content on material extrusion (MEX) 3D-printed composites using acrylonitrile butadiene styrene (ABS) as the polymer matrix and 0–10 vol% hexagonal boron nitride (BN) as reinforcement. ABS/BN composites were prepared through batchwise compounding and extruded into MEX-suitable filaments. The filaments were conditioned at 30 °C and 10% or 90% relative humidity (RH) before/during direct feeding into the 3D printer. Specimens were fabricated with raster angles parallel (0°) and perpendicular (90°) to their length. Micro- and macrostructural analyses using scanning electron microscopy and computed tomography revealed intensive void formation, especially in BN-filled composites 3D-printed from humid filaments. This was attributed to BN acting as a physical barrier, hindering the outgassing of evaporated water during 3D printing. Mechanical properties were evaluated using tensile and Charpy impact tests. Based on the tensile test results, neat ABS was the least sensitive to filament moisture, with tensile strength at 0° raster angle dropping from 40.5 MPa to 36.7 MPa as storage RH was increased from 10 to 90%. For composites with 10 vol% BN loading, tensile strength dropped from 34.1 MPa to 22.3 MPa. Charpy impact strength exhibited similar reductions, ascribed to the porous structure of the BN-filled composites caused by the evaporated moisture. Thermal conductivity was also examined, showing slightly superior performance for samples 3D-printed from filaments stored in less humid conditions. For unfilled ABS, the conductivity slightly decreased from 0.188 to 0.185 W/mK, while for 10 vol% BN-filled composite, it dropped from 0.778 to 0.617 W/mK.

Material extrusion (MEX), also commonly referred to as fused deposition modeling (FDM) or fused filament fabrication (FFF) is currently one of the most commonly used additive manufacturing techniques. The quality of the 3D-printed objects fabricated by MEX methods highly relies on various printing parameters, one of which is the so-called filament extrusion multiplier (k). In this study, 3D-printed parts were prepared by MEX technique during which the material feeding rate was adjusted by varying the extrusion multiplier in the range of 97–105% (k = 0.97–1.05). The fabricated parts were tested for their geometrical, structural, mechanical, and thermal conductivity properties. Based on computed tomographic analysis and scanning electron microscopic images, increasing the k parameter resulted in smaller voids, along with gradually decreasing porosity (from 5.82 to 0.05%). Parallel to the decreasing defects, the thermal conductivity of the parts improved from 0.157 to 0.188 W/mK as determined by light-flash analysis technique. On the other hand, when k was set to ≥ 1.03 the geometrical accuracy declined, the size of the specimens considerably increased relative to the nominal values, especially in the X–Y directions due to excess material getting “squeezed” on the sides of the specimens. This latter phenomenon also resulted in the formation of a number of stress concentration sites, which manifested in the decrease of mechanical properties. Accordingly, the tensile, flexural, and impact strength of the samples improved up to k = 1.03; however, above that it dropped considerably.

The purpose of this work is to propose a decision-making algorithm to select the optimal composite material for thermally conductive but electrically insulating applications, such as microelectronic packaging heat sinks, diodes, and other electronic devices. In particular, an algorithm based on the criteria importance through inter-criteria correlation (CRITIC) and additive ratio assessment (ARAS) methods are used to evaluate several conflicting attributes. The evaluated samples were acrylonitrile butadiene styrene (ABS) composites filled with 0–30 vol% of boron nitride (BN) particles and prepared through melt compounding. The performance attributes considered through testing were heat conductivity, electrical resistivity, density, hardness, and tensile properties (Young's modulus, tensile strength, and elongation). As expected, the composite containing 30 vol% BN exhibited the highest heat conductivity, electrical resistivity, and Young's modulus. Meanwhile, unfilled ABS had the highest elongation at break, tensile strength, and lowest density. With respect to hardness, the 1 vol% BN-loaded composite proved to be superior. Therefore, the experimental data revealed a considerable compositional dependence with no obvious trend. The optimal composition was identified by adopting the CRITIC-ARAS multi-criteria decision-making algorithm, based on which the 30 vol% BN-containing composite was dominant among all the prepared samples. A validation through other decision-making techniques was performed to support the robustness of the proposed technique. Additionally, a sensitivity analysis was carried out on several weight exchange scenarios to see the stability of the ranking results.