David Safranski1 Andrew Miller2 Catherine Wood2 Robert Guldberg3 Ken Gall2

1, MedShape, Inc., Atlanta, Georgia, United States
2, Duke University, Durham, North Carolina, United States
3, Georgia Institute of Technology, Atlanta, Georgia, United States

Polyurethane (PU) based elastomers continue to gain popularity in a variety of biomedical applications due to their low stiffness, favorable biocompatibility, and high strength. In parallel, advancements in additive manufacturing continue to provide new opportunities for biomedical applications by enabling the creation of more complex architectures for tissue scaffolding and patient specific implants. Optimizing implants for success in fatigue-prone applications depends on a strong understanding of the relationship between material structure and fatigue performance, a surprisingly understudied area.
First, we sought to develop relationships between PU structure and mechanical properties, including fatigue, for three soft PUs with systematically varied ratios of hard and soft segments. In addition, we compared injection molded controls to 3D printed (fused deposition modeling, FDM) varieties to examine the effects of such processing. Second, we examined the effects of printed architecture on the monotonic and cyclic mechanical behavior of elastomeric PUs and to compare the structure-property relationship across two different printing approaches. We examined the tensile fatigue of notched specimens, 3D crosshatch scaffolds, and two 3D spherical pore architectures in a physically crosslinked polyurethane printed via FDM as well as a photo-cured, chemically-crosslinked, elastomeric PU printed via continuous liquid interface production (CLIP).
Results indicate that increased hard segment content leads to increased stiffness, increased shear failure stress, and improvements in tensile fatigue from a stress-based standpoint despite relatively uniform tensile strength for the tested grades. Effects of hard segment content on tensile failure strain, and strain-based fatigue performance, were more complex and largely influenced by microphase organization and interaction. FDM samples matched or exceeded injection molded controls in terms of tensile failure stress and strain, compressive properties, shear strength, and tensile fatigue. The success of FDM samples is attributed in part to favorable printing parameters and the toughness of PU which results in lower flaw sensitivity. PUs from both FDM and CLIP were relatively tolerant of 3D geometrical features as compared to stiffer synthetic implant materials such as PEEK and titanium. PU and crosslinked PU samples with 3D porous structures demonstrated a reduced tensile failure stress as expected without a significant effect on tensile failure strain. PU crosshatch samples demonstrated similar performance in strain-based tensile fatigue as solid controls; however, when plotted against stress amplitude and adjusted by porosity, it was clear that the architecture had an impact on performance. Square shaped notches or pores in crosslinked PU appeared to have a modest effect on strain-based tensile fatigue while circular shaped notches and pores had little impact relative to smooth samples.