Computational Methods in the Modeling of Scaffolds for Tissue Engineering - Computational Modeling in Tissue Engineering - page 113

Fig. 7 Representative threshold z-stacks of confocal microscopy images (500 lm thickness)
showing adhering cells after perfusion seeding. The fluid shear rates were measured in the
scaffold surface through the CFD methods. a Type I. b Type G (from Melchels et al. [ 12 ])
scaffold (type I) were of 412 ± 13 lm with porosity of 62 ± 1% (Fig. 7 a). In
order to study the influence of pore size variation, a gradual variation of pore size
(type G) was introduced where the scaffold pore size was &500 lm at the center
and &250 lm at the periphery with a total porosity of 56 ± 3% [ 12 ]. Fluid
volume was reconstructed from micro-CT images and represented the actual
design of the perfusion chamber. These models assumed steady state conditions for
maximum fluid flow while an alternant flow was applied in vitro when the cells are
suspended in the medium. However, even with this restriction, the computational
model demonstrated a close relation between distribution of maximum stimuli and
final cell seeding distribution (Fig. 7 ).
When the distribution of pore size was uniform (Type I), the distribution of
cells densities observed through confocal images after seeding was also homo-
geneous, in a similar behavior obtained numerically (Fig. 7 a). In the gradient pore
size (Type G), a gradient of shear stress was obtained in the cylindrical scaffold,
which compared well with the gradual distribution of cells obtained after 16 h of
seeding (Fig. 7 b).