Page 288 - IJB-9-4
P. 288

International Journal of Bioprinting                 3D printing of continuous fiber reinforced PLA/PGA composites

















































            Figure 9. 3D full-field strain measurement system (3DFFSMS) analysis results. (A) Scheme 3, (B) specimen of scheme 3, (C) scheme 4, (D) scheme 6, (E)
            specimen of scheme 4, and (F) failed specimen of scheme 6.

            area, and the area of the red region gradually expanded   the specimen of scheme 5 was similar to that of scheme 3
            until the specimen fractured. As shown in Figure 9E, the   shown in Figure 9A.
            fracture location of the specimen was consistent with the
            location of the red stress concentration area at point V on   3.3. Results of internal defect analysis
            the stress-strain curve. Unlike the specimen of scheme   The presence of internal porosity defects between
            3, the bonding between the PLA matrix and PGA fibers   deposition lines is a common problem in FDM printing.
            was good, and no debonding situation occurred until the   These internal defects are related to the printing process
            specimen failed.                                   parameters,  which  will  affect  the  performance  of  a
                                                               printed part. A specimen was randomly selected from
               The tensile process of the specimen of scheme 6 was   each experimental scheme for internal defect analysis.
            similar to that of scheme 4. The stress–strain curves, strain   Figure 10 shows the analysis results of the internal defects
            cloud and fractured specimen of scheme 6 are illustrated in   of the specimens obtained by the micro-X-ray 3D imaging
            Figures 9D and F. In the strain cloud history, the specimen   system. Different colors were used to represent the sizes
            showed lower strain values than that of scheme 4, which   of  the  internal  voids  in  the  figure,  and  the  measured
            means that the tensile strength of the specimen of scheme   porosities are shown in  Table 2. The internal defects of
            6 was inferior to that of scheme 4. The shapes of the stress-  specimens printed with different process parameters were
            strain curves and post-fracture morphologies of specimens   also different. The defects in scheme 3 (Figure 10) were
            of schemes 7 and 8 were similar to those of scheme 6   clearly visible, and there were more defects in the upper
            shown in Figure 9D. The shape of the stress–strain curve of   layer than in the lower layer.


            Volume 9 Issue 4 (2023)                        280                         https://doi.org/10.18063/ijb.734
   283   284   285   286   287   288   289   290   291   292   293