To stably extrude molten filament, Pgas is an important parameter that determines the smooth deposition of molten filament and keeps Qgas unchanged. As shown in Fig. 4(a), when Pgas is lower than the extrusion melt pressure (Pmelt), the effect of swelling away from the mold is not eliminated, gas is compressed by extrusion of the melt at the beginning, resulting in a smaller gas flow, gas stuck in nozzle could only exit gradually, until the pressure is increased to a certain value, a break-through flow causes all the gas to be released, resulting in uneven continuous extrusion of the surface of the wire, thus forming one bamboo node after another. As shown in Fig. 4(c), when Pgas exceeds the melt extrusion pressure (Pmelt), strong airflow occupies the whole nozzle export channel, the melt is not subject to gas compression, continuous accumulation in the nozzle exit causes the melt pressure to increase to a certain value, one-time accumulation of all melt extrusion results in continuous extrusion of the surface of the filament being incomplete, thus forming one “lump” after another. As shown in Fig. 4(b), by setting Pgas equal to the melt pressure (Pmelt), the state of gas and melt is relatively stable. At this time, a continuous thin layer of lubricating gas is formed on the inner wall of the nozzle, and smooth deposition of FDM filament is realized. According to the analysis of the experimental results, when Pgas = 0.4 MPa, a stable gas-assisted layer is formed on the inner wall of the nozzle, and smooth extrusion and deposition of FDM wire can be realized. No extrusion swelling phenomenon occurs, and the expected result is achieved. Fig. 4(d) shows that when Pgas = 0.4 MPa, the surface of the printed parts is the smoothest and the quality thereof is optimized. To form a stable gas-assisted layer on the inner wall of the nozzle, Qgas is also an important parameter that must be controlled, and Pgas remains unchanged at 0.40 MPa at this time. As shown in Fig. 4(e), when Qgas < 1.75 L/min, the inner wall of the nozzle cannot form a stable gas thin layer of gas lubrication. When the melt is extruded, the high shear and surface tension caused by the micro-size effect make the effect of swelling away from the mold more obvious. This results in the uneven surface of the continuous extruded filament, thus forming one gap after another. As shown in Fig. 4(g), when Qgas > 1.75 L/min, the gas occupies almost the entire outlet channel and the melt diameter is compressed to less than 1 mm, resulting in uneven surface and small diameter of continuous extruded filament. As illustrated in Fig. 4(f), when Qgas = 1.75 L/min, a stable thin layer of gas lubrication is formed on the inner wall of the nozzle, eliminating swelling away from the mold and resulting in smooth deposition of filament. Fig. 4(h) shows that when Qgas = 1.75 L/min, the surface smoothness of the printed parts is maximized and the molding quality is optimal (this also matches the expected results).
As shown in Figs. 5(a), (b), and (c), at a given Qgas, the tensile strength first increases, then decreases with the increase of Pgas, and reaches the peak value when Pgas is 0.4 MPa. It is found that when Pgas is less than 0.4 MPa, the surface of the extruded filament is not uniform. Although the interlayer temperature and molecular diffusion time do not change, the gap between the deposited filament increases and the relative layers adhesion area is small, resulting in a low tensile strength. As Pgas approaches 0.4 MPa, the surface of the extruded filament is almost smooth and complete, and the sample-forming quality is improved, so the tensile strength continues to increase to its maximum value. As Pgas begins to move away from 0.4 MPa, the gap between the deposited filament begins to increase, the relative adhesion area becomes smaller, and the phenomenon of stress concentration appears at the fracture, leading to the decrease of tensile strength and the decrease of the sample-forming quality.
As shown in Fig. 5(d), the increment of tensile strength of the printed parts with gas-assistance is compared with that without gas-assistance. The result indicates that the rate of increase in tensile strength is the highest when Qgas is 1.75 L/min and Pgas is 0.4 MPa, while it is the lowest when Qgas is 2 L/min and Pgas is 0.45 MPa (albeit it remains above 50% of its maximum value). The enhanced mechanical strength obtained by printing parts with gas-assisted may be a combination of several factors. Firstly, the instantaneous heating and pressurizing effects of the high-pressure hot airflow on the surface layer of printing are conducive to the interlayer macromolecular chain infiltration, diffusion and entanglement. Secondly, the decrease of porosity in gas-assisted printed parts can enhance the inter-layer adhesion. Finally, the factor that may increase the tensile strength of the printed part is the reduction of residual stress. It has been widely reported that, at standard atmospheric pressure, uneven thermal gradients and stress accumulation result from repeated heating and cooling cycles during deposition of the upper layer of print. In the case of input high-pressure hot airflow deposition, since the heat loss is very small and the printed part is not completely cooled, the manufactured parts are unaffected by the same thermal gradient or cooling cycle. In addition, the reduction in cooling of prints printed under gas-assisted conditions can enhance the inter-layer adhesion by thermally driven molecular diffusion between layers, thus improving the tensile strength of the printed parts.
To ascertain the reasons for the improvement of mechanical properties of gas-assisted printed parts, an SEM (SU1510) is used to scan the tensile cross-section of the samples to observe the interlayer adhesion of the samples (Fig. 6). As shown in Fig. 6(a), at Qgas = 1.75 L/min and Pgas = 0.30 MPa, obvious voids and cracks appear within the section. This is because Pgas is too small. When the gas is accumulated to equal to Pmelt, the diameter of the extruded filament becomes smaller when the gas is flushed out in one rapid event, and the surface of the filament is uneven, showing one bamboo node after another. There are gaps between the adjacent filament on the printing surface, leading to poor mechanical properties. As shown in Fig. 6(b), with the increase of gas pressure, when Qgas = 1.75 L/min and Pgas = 0.35 MPa, the void within the section disappears, the crack is ameliorated, there are obvious brittle fracture zones, and the tensile strength is improved, which is consistent with the measured mechanical properties. As shown in Fig. 6(c), at Qgas = 1.75 L/min and Pgas = 0.40 MPa, the surface of the fracture is smooth, the voids are practically eliminated, and the interlayer quality and adhesion are optimized. As shown in Fig. 6(d), at Qgas = 1.75 L/min and Pgas = 0.45 MPa, cracks begin to worsen, voids expand, and defects and stress concentration are formed in the accumulation of filament. This is because Pgas is too large. When the accumulation of Pmelt exceeds Pgas, the melt breaks out in a single event, leading to the accumulation of filament. Stress concentration ensues, resulting in poor mechanical properties.
As shown in Fig. 7, to assess the effect of the gas-assisted nozzle printing example model designed in this study, three sample models were printed under the optimal gas parameters, Qgas = 1.75 L/min and Pgas = 0.4 MPa. For thin-walled parts (Figs. 7a and 7d), compared with the conditions without gas-assistance, the surface of the air-assisted printed parts is found to be smooth from the surface, and the printing track is clearly observed. From the edge, the edge of the printed parts shows no excess burring and the surface is smooth, and the fusion between layers is ideal, with almost no gap. For cuboidal parts (Figs. 7b and 7e), compared with the conditions without gas-assistance, the gas-assisted printed parts have distinct connections at each surface, without excess burr and debris, exhibiting good molding effects. For columnar parts (Figs. 7c and 7f), compared with the conditions without gas-assistance, the surface of the air-assisted printed parts fits well with the side connection, without excess burrs, and the side surface has no obvious boundary, no obvious warping and stratification, with uniform texture and good molding effects. Fig. 7(g) shows the rate of shrinkage in dimensional accuracy of the printed sample model at Qgas = 1.75 L/min and Pgas = 0.4 MPa without gas-assistance. In general, compared with the non-gas-assisted condition, the shrinkage of the parts under gas-assisted condition in the three directions of length, width, and height decreases to varying degrees, with the greatest shrinkage being in the height direction. At the same time, the size error shrinkage rate in the three directions of the gas-assisted printing parts is within 2.5%, and the accuracy is relatively high. Among them, the height direction of the columnar parts is the best (the shrinkage rate therein is only 0.13%), therefore, the design of the print head based on the combination of the proposed gas-assistance and FDM printing technique improves the dimensional accuracy and mechanical strength between layers of 3D printed parts, and promotes the application of 3D printing in medical, aerospace, automotive, and other fields.