Search

In this study, we compared the material properties of linearly and sharply graded Ti6Al4V additively manufactured samples to investigate whether the more severe discontinuities caused by sharp grading can reduce performance. We performed compression testing with digital image correlation (DIC) in two loading directions for each grading design to simulate iso-stress and iso-strain conditions. We extracted the elastic stiffness, yield strength, yield strain, and energy absorption capacity of each sample. In addition, we used micro-computed tomography (micro-CT) imaging to examine the printing quality and dimensional accuracy. We found that sharply graded struts have a 12.95% increase in strut cross-sectional areas, whereas linearly graded struts produced an average of 49.24% increase compared to design. However, sharply graded and linearly graded FGL samples do not have statistically significant differences in elastic stiffness and yield strength. For the iso-strain condition, the average DIC-corrected stiffnesses for linearly and sharply graded samples were 6.15 GPa and 5.43 GPa, respectively ( p = 0.4466), and the yield stresses were 290.4 MPa and 291.2 MPa, respectively ( p = 0.5734). Furthermore, we confirmed different types of printing defects using micro-CT, including defective pores and disconnected struts. These results suggest that the loss of material properties caused by manufacturing defects outweighs the adverse effects of discrete-grading-induced discontinuities.

Endoprosthetic reconstruction of the pelvic bone using 3D-printed, custom-made implants has delivered early load-bearing ability and good functional outcomes in the short term to individuals with pelvic sarcoma. However, excessive stress-shielding and subsequent resorption of peri‑prosthetic bone can imperil the long-term stability of such implants. To evaluate the stress-shielding performance of pelvic prostheses, we developed a sequential modeling scheme using subject-specific finite element models of the pelvic bone-implant complex and personalized neuromusculoskeletal models for pre- and post-surgery walking. A new topology optimization approach is introduced for the stress-shielding resistant (SSR) design of custom pelvic prostheses, which uses 3D-printable porous lattice structures. The SSR optimization was applied to a typical pelvic prosthesis to reconstruct a type II+III bone resection. The stress-shielding performance of the optimized implant based on the SSR approach was compared against the conventional optimization. The volume of the peri‑prosthetic bone predicted to undergo resorption post-surgery decreased from 44 to 18%. This improvement in stress-shielding resistance was achieved without compromising the structural integrity of the prosthesis. The SSR design approach has the potential to improve the long-term stability of custom-made pelvic prostheses., Competing Interests: Declaration of Competing Interest The authors state no conflict of interest for the study object of the manuscript., (Copyright © 2023. Published by Elsevier Ltd.)

The orthopedic industry is still searching for an efficient way to replace bone loss due to surgical procedures such as arthroplasty and limb-sparing surgery. Additive manufacturing (AM) presents an opportunity to manufacture affordable patient-specific implants. Optimization of the implant-bone interface to maximize osseointegration (bone ingrowth) has not been appropriately addressed. Mechanobiological models, suited to predict mechanical adaptation of bone, cannot be used to predict osseointegration inside implants as the implant is not exposed to any mechanical loading until it is fully accepted by the host body. Biological models relying on partial differential equations based on continuum approximation are not well-suited to predict the discrete phenomenon of osseointegration. This study proposes an agent-based modeling (ABM) approach for representing the osseointegration process for orthopedic implants produced by powder-bed additive manufacturing processes. Agent-Based Modeling (ABM) is a cellular automata based discrete computing technique that uses rule-based mathematics derived from experimental studies to simulate evolutionary phenomena. In this paper, osseointegration inside a hexagonal closed packing of AM powder particles is modeled using ABM. Cellular agents such as pre-osteoblasts and osteoblasts are realistically modeled as cubic cells. The proposed model underpredicts osseointegration at early stages but predicts osseointegration at around 21 days with sufficient accuracy when compared to the in vitro test conducted by Xue et al. in 2007., (Copyright © 2021. Published by Elsevier Ltd.)

Patient-specific finite element (FE) models of bone require the assignment of heterogeneous material properties extracted from the subject's computed tomography (CT) images. Though node-based (NB) and element-based (EB) material mapping methods (MMMs) have been proposed, the sensitivity and convergence of FE models to MMM for varying mesh sizes are not well understood. In this work, CT-derived and synthetic bone material data were used to evaluate the effect of MMM on results from FE analyses. Pelvic trabecular bone data was extracted from CT images of six subjects, while synthetic data were created to resemble trabecular bone properties. The numerical convergence of FE bone models using different MMMs were evaluated for strain energy, von-Mises stress, and strain. NB and EB MMMs both demonstrated good convergence regarding total strain energy, with the EB method having a slight edge over the NB. However, at the local level (e.g., maximum stress and strain), FE results were sensitive to the field type, MMM, and the FE mesh size. The EB method exhibited superior performance in finer meshes relative to the voxel size. The NB method converged better than did the EB method for coarser meshes. These findings may lead to higher-fidelity patient-specific FE bone models., (Copyright © 2021. Published by Elsevier Ltd.)