Page 384 - IJB-10-2

International Journal of Bioprinting AM evaluation of medical device companies

potentially risked implant performance. It is important to fatigue strength than heterogeneous morphologies and
20
14
highlight that all the companies participating in the study could have been a consequence of correctly applying thermal
are already manufacturing implants that have been used in methods such as HIP.
12
clinical practice. Most of these companies are ISO 13485 The rest of the prototypes present heterogeneous grain
certified, demonstrating that this compliance does not morphology with dendritic distributions. P3 could be
guarantee the high quality of the manufacturing. next in the ranking as it shows the most homogeneous
P3 and P7 showed maximum deviations of 5 mm. microstructure after P5 and the porosity is of spherical
P7 was delivered without the entire L2 and without a morphology. P4 could be third as it exhibits no defects in
29
warning or feedback about the issue. P1, P3, P7, and P8 the surface and low content of porosity. P1 demonstrated
30
were provided with broken lattice structures, and only C1 some discontinuities in the surface and in the matrix, but
communicated the issue when shipping the part. Only the grain morphology was more favorable than P6, P2, and
P4, P5, and P6 produced L2 correctly. This demonstrates P8, that showed very heterogeneous grain morphology and
that the production workflow and printing parameters thinner grains than the rest of prototypes, indicating the
can determine whether a design feature is feasible to least favorable mechanical responses. P6 and P2 (similarly
26
be manufactured or not. Manufacturers should assess manufactured by C2) presented similarities with some
whether they are capable of manufacturing a design with very large inclusions. The latter was more favorable as it
31
their resources and expertise, and identify potential design presented less inclusions in the matrix. P8 could be next
issues for optimal printing, applying DfAM practices. as it showed a visible layer of alpha-case, which indicates
In this study, only two companies, C4 and C5, provided oxidation and corrosion. This layer can cause a negative
feedback about the design before fabrication to optimize effect on the mechanical properties of the part such as
it for AM. This highlights the need for improvement ductility, fracture toughness, and fatigue life, especially
in DfAM habits among manufacturers. It is unknown under dynamic loading. 25
whether the production issue for L2 was caused during P7 presented the most harmful defect for the integrity
printing or postprocessing, but the L2 architecture with of the part. Although the microstructure had favorably
a minimum beam thickness of 0.5 mm was confirmed to large grains, the part showed an internal crack originating
be achievable by all the companies prior to production. in the surface that extended toward the center of the
Another nonconformity was found in P2, which had a matrix, potentially leading to early failure. Furthermore,
hole 2 mm bigger in diameter and 2 mm out of the correct the component was delivered with the entire surface of
location. Interestingly, P6 was manufactured by the same lattice L2 missing, registering deviations of up to 5 mm. It
company, but did not present any geometrical deviations, was also sent with a matte finish, although a mirror finish
showing the inconsistency of the same company in the had been agreed before production. Yet, C6, the company
quality delivered.
that produced P7, is legally certified to commercialize class
The microstructural analysis also showed some III custom implants and complies with the ISO 13485.
deficiencies such as large pores and cracks in the surfaces, It is important to find a balance between postprocessing,
mainly in P8 but also in P1 and P2, that could potentially costs, and functionality during the production of a
trigger early fractures. Different grain morphologies specific implant. Thermal postprocesses can improve
were also found between the prototypes. We estimated a mechanical performance but also have an effect on
potential ranking of microstructures, from the best to the manufacturing costs. The optimal microstructure required
worst, as follows: P5, P3, P4, P1, P6, P2, P8, and P7. for a maxillofacial cranial plate may be different than for
P5 demonstrated to be the most superior prototype for the a pelvic reconstruction as the former is not subjected to
particular function of the pelvic reconstruction. This is revealed high fatigue loading, as opposed to the latter. Therefore,
in the accuracy of the geometry, the homogeneity of the implant they may need different production steps. With the lack
surfaces, the low content of porosity in the sample analyzed, of regulatory and standardization guidelines to develop
and the globular morphology of the grain and homogeneous specific implants with AM, engineers must think critically
alpha and beta phases. Metal AM parts, especially in PBF, about what production process is required according
usually show columnar-oriented microstructures. The to the biomechanical requirements of the patient. The
3
production of P5 achieved recrystallization and transformed different implant requirements are going to determine the
the columnar microstructure to equiaxed microstructure characterization of the material and manufacturing steps
with homogeneous and globular small equiaxed grains. This to achieve a balance between microstructure, mechanical
microstructure has been demonstrated to provide better properties, and cost for each individual application.

Volume 10 Issue 2 (2024) 376 doi: 10.36922/ijb.0140