Page 219 - IJB-10-2
P. 219

International Journal of Bioprinting                                Property of scaffolds with different lattices




            Table 2. Parameters of the model                   was exported. Ti6Al4V material properties (Poisson’s ratio
                                                               of 0.376 and elastic modulus of 110 GPa) were assigned
             Experimental   Type of lattice  Partition   Number of
             type                    grid type   grids (10 )   to the model after the saved files were imported into
                                                      4
             Compression    CPL        C3D4        619         the Abaqus software. The bottom end of the model was
                                                               held in place, while the higher end was subjected to the
             test         Diamond      C3D4        835         displacement load. The  stress–strain curve  was plotted
                        Cuboctahedron  C3D4        839         based  on  the  measured  data.  According  to  the  slope  of
             CFD            CPL        C3D4        666         the  elastic  deformation  zone  in  the  stress–strain  curve,
                          Diamond      C3D4        873         we estimated the elastic modulus of the scaffold, drew
                        Cuboctahedron  C3D4       1096         a straight line using the elastic deformation stage, and
                                                               then translated the line by 0.2%. The intersection point
                                                               obtained was the yield strength. In order to test the actual
            and measured the real pore size and rod diameter using   mechanical properties of titanium alloy porous scaffolds
            scanning electron microscopy (SEM). Simultaneously, the   with different lattice structures, mechanical compression
            scaffolds were scanned with micro-computed tomography   tests were carried out in vitro. During the experiment, the
            (micro-CT; voltage 155 kV, current 120 µA, resolution 17   porous scaffolds with different lattice structures (diameter
            µm,  projection  number  1440,  integration  time  500  ms).
            Then, the VG Studio MAX3.5 software (Volume Graphics,   10 mm, height 10 mm) were placed on the universal test
            Germany) was employed to redevelop the 3D structure of   machine (C43.104, MTS Ltd., China) to make the long
            the scaffold, and the internal structure of the scaffold was   axis of the scaffolds consistent with the direction of the
            analyzed. The real porosity of the porous Ti6Al4V scaffolds   application force (Figure 3B). The load device drops at a
            was  determined using  the  dry  weight  method.  In  other   constant speed of 1 mm/min. According to the software
            words, under typical air pressure, the porous scaffolds   equipped with the loading system, the stress–strain curve
            of each group were in dry state. The actual porosity of   was drawn (Figure 3C), and the yield strength and elastic
            the scaffolds was computed using the following formulas   modulus of the scaffolds were calculated (Table 4).
            once the actual weight of the scaffolds was determined   2.4. Computational fluid dynamics analysis
            (Equations VII and VIII):                          Utilizing computational fluid dynamics (CFD), the
                                    m                          hydrodynamic properties of porous scaffolds with different
                                  V 0               (VII)    unit cell types were analyzed. The scaffold model (Φ10
                                 0
                                                               mm × 3 mm) was a hard, inflexible body (see Table 2 for
                                                               specific parameters). The fluid used was an incompressible
                                                            and homogeneous tissue fluid with a temperature of 37°, a
                              P  1   0            (VIII)

                                                                                           −3
                                                            viscosity coefficient (η) of 3.20 × 10 Pa·s, and a density (ρ)
                                                                          3
                                                               of 1060 kg/m . In order to avoid the influence of the fluid
            where ρ  represents the apparent density (g/cm ), and m   domain boundary on the experimental structure, the fluid
                                                   3
                  0
            denotes the actual weight (g) of the porous titanium alloy   domain size was Φ10 mm × 5 mm, and the porous model
            scaffold.  V  denotes the volume of the porous titanium   was positioned at the center of the fluid domain model. We
                     0
            alloy scaffold in its natural state (cm ), P shows the actual   defined the fluid inlet surface, the outlet surface, and the
                                         3
            porosity of the porous titanium alloy scaffold, and  ρ   inner wall of the scaffold (Figure 3D). The fluid domain
            represents the  theoretical density of  Ti6Al4V  material    and rigid surface satisfied the no-slip condition, and the
                   3
            (4.5 g/cm ).
                                                               normal and tangential vectors at the fluid–solid interface
            2.3. Compressive mechanical testing                were both zero. The fluid entered the fluid channel at a
            Static simulation analysis was employed to test the   flow rate of 0.1 mm/s from the intake surface. The outlet
            compressive  mechanical  properties  of  porous  Ti6Al4V   surface of the fluid channel is the fluid outlet, which is
            scaffolds with different unit cell structures (Φ10 mm  ×   defined as the free boundary condition, i.e., the pressure
            10 mm). Importing 3D models of scaffolds with various   is 0 MPa. We investigated the specific surface area of the
            unit cell configurations into the Hypermesh program and   scaffold (Equation IX) in addition to the fluid flow velocity
            using the surface deviation function to mesh the models   through the porous scaffold and the shear stress created
            is a unique procedure (Figure 3A). The mesh parameters   on the scaffold’s inner wall. In addition, the permeability
            are illustrated in Table 2. After using the cleanup function   of the scaffold was determined using the Darcy formula
            and cleanup tools to repair the mesh, the .inp format file   (Equation X).
                                                                          43



            Volume 10 Issue 2 (2024)                       211                                doi: 10.36922/ijb.1698
   214   215   216   217   218   219   220   221   222   223   224