Page 91 - IJB-6-2

Kolan, et al.
Table 2. Scaffold porosity and pore size.

Architecture Designed versus apparent (%) Measured average pore size (µm)
50 60 70 80 50 60 70 80
Cubic 33±2 40±2 54±1 66±3 0.5±0.02 0.7±0.02 1.0±0.04 1.3±0.02
Spherical 32±1 42±3 49±3 61±1 0.5±0.04 0.6±0.05 1.0±0.04 1.0±0.04
X 28±2 35±3 43±2 55±0 0.5±0.03 0.6±0.02 0.7±0.03 1.2±0.10
Diamond 34±2 40±1 47±1 61±1 0.8±0.01 0.9±0.01 1.0±0.04 1.2±0.05
Gyroid 34±1 41±1 49±1 60±1 0.8±0.01 0.9±0.01 1.0±0.03 1.1±0.04

between 0.5 and 1.3 mm. X architecture scaffolds subtracted from the solid unit cube to obtain the
had the smallest pores which are consistent with unit cell of spherical architecture. Figure 2A
the X scaffolds having the largest difference shows unit cells of spherical and cubic scaffolds
between the designed and apparent porosities. and their pore shapes and pore sizes. Decreasing
The difficulty in removing adhered powder from sphere diameter to lower porosity would hinder
the X scaffold pores of green bodies contributed the removal of powder particles from the scaffold.
to its lowest porosity and most reduced pore size. Hence, unit cell pores were designed to enable
The average pore size of scaffolds designed with fabrication and removal of powder particles. The
80% porosity ranged from 0.9 mm to 1.3 mm, and pore volume variation for cubic and spherical unit
the scaffolds designed with 50% porosity ranged cells is shown in Figure 2B.
from 0.5 mm to 0.8 mm. In theory, unit cell pore
size varies along the Z-height, with it being the 3.2 Effect of porosity and pore geometry on
smallest at the end and the largest at the center of mechanical strength
the unit cell. Therefore, true scaffold pore size has Compressive strengths of borate glass scaffolds
a range of values instead of being a constant value. with different porosities are shown in Figure 3A.
Several parameters limit the accuracy
of fabricating scaffolds, including scaffold Among the five architectures investigated in this
architecture, the resolution of the machine, layer study, cubic scaffolds had the highest compressive
thickness, binder content, and particle size. The strength (15.5 ± 1.9 MPa) and X scaffolds had the
laser spot diameter of the SLS machine was lowest strength (4.9 ± 1.2 MPa) at low porosity
0.45 mm and therefore, it was not feasible to (~35%). The biomimetic architectures (gyroid and
fabricate scaffolds with struts smaller than this diamond scaffolds) had compressive strengths of
limit (<0.45 mm). The laser spot could have 9.5 ± 2.5 MPa and 6.8 ± 1.6 MPa, respectively.
heated and melted particles adjacent to the The scaffold compressive strengths at low
scanning area, effectively reducing the designed porosity were near the high end of the range of
pore size. Smaller particles are easier to remove compressive strength for human trabecular bone
from the green body scaffold pores, causing (~2 – ~12 MPa), whereas the strengths at high
less deviation from the actual design. However, porosity levels (>55%) were near the low end of
smaller particles require higher binder content the spectrum [39] . Cubic architecture scaffolds have
because of the increased surface area that could pillars in the axial direction that carry a majority
increase the shrinkage and deviations between of the load in compression tests before structural
designed and fabricated parts. The amount of failure, while the other architectures lacked a
binder and the particle size was optimized for similar feature. X architecture scaffolds provided
scaffold fabrication in our previous work . the least resistance in compression because of
[35]
One key aspect in designing the architecture the 45° oriented struts. The compressive strength
was considering scaffold manufacturability. for all scaffold types was ~4 MPa or less at
For instance, in spherical scaffolds, porosity is high porosities, which falls at the lower range
a function of the diameter of the sphere that is of the trabecular bone compressive strength [39] .

International Journal of Bioprinting (2020)–Volume 6, Issue 2 87