Computed tomography evaluation of the porosity and fiber orientation in a short carbon fiber material extrusion filament and part

The computed tomography (CT) evaluation of the material extrusion (MEX) of a short carbon fiber (SCF) Nylon-12 filament and part is presented. CT, a non-destructive testing method, was used to quantify the internal structure of specimens into three phases: pore, Nylon, and SCF. The intensity histograms from the CT data were fit using a mixed skew-Gaussian distribution (MSGD) algorithm to segment the CT image into phases. Thresholded images were used to isolate pores in the CT image to determine pore volume and distribution within both the MEX SCF filament and part. The phase volume percentages of the MEX SCF filament were found to be 1.6% pore, 62.2% Nylon, and 36.2% SCF. The volume of most pores within the filament were found to be under 100 μm3. The highest frequency of pores was located near the outside of the filament, but the large pores were located near the center of the filament. This result indicates that the thermoplastic filament extrusion process likely entraps large bubbles in the center of filament or causes large thermal gradients and residual stresses that induce voids during post-extrusion cooling. MSGD analysis of sections of the MEX SCF part estimated phase volume percentages to be 9.8% pore, 59.6% Nylon, and 30.9% SCF. This analysis showed a more than 8% increase in porosity from the MEX SCF filament to the part. For the MEX SCF part, the average pore area was found to be highest (>250 μm2) at the bottom of the layer and smallest (<100 μm2) at the top of the layer, which could be explained by a large temperature gradient between and contractile thermal stresses inside the layer that cause the thermoplastic to shrink into a smaller volume allowing the voids to grow during deposition. A qualitative analysis of fiber orientation conducted on the SCF filament indicated that the SCFs maintain their orientation from filament to part except in the intersection zone of rasters.