David Safranski1 Nathan Evans2 Cameron Irvin2 Cambre Kelly3 Ken Gall3

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

Additive manufacturing of Ti-6Al-4V ELI is becoming a leading method of producing orthopaedic devices with complex geometries and for patient–specific implants. Selective laser melting (SLM) is one of the leading choices for powder based additive manufacturing. The fatigue properties of SLM Ti-6Al-4V have been studied for a variety of processing conditions and geometries. However, these tests had varying heat treatments, geometries, R-values, and frequencies. The purpose of this work is to systematically determine the effect of surface treatments and varying structure on the fatigue behavior of SLM Ti-6Al-4V ELI. The monotonic tensile and fatigue behaviors were examined to determine the relationships between topography, porosity, and mechanical performance.
Dogbone test specimens were manufactured by selective laser melting of Ti-6Al-4V ELI (3D Systems DMP320). Three specimen geometries were built: solid, solid with an additional 0.5mm porous layer on all sides, and a 65% porous gage section. The porous structure was based upon a diamond lattice. All samples underwent a hot isostatic press. Then, samples underwent one post-processing treatment: as-built (no treatment), rotopolishing, or SILC cleaning. All samples were subsequently anodized according to SAE AMS 2488D Type 2. Samples were tensile tested to failure at a displacement rate of 1 mm/min using a MTS Satec 20 kip servo-controlled, hydraulically-actuated test frame. Fatigue tests were run at increasingly lower stress levels below the yield strength of the samples to generate fatigue curves and to determine the endurance limit. Fatigue tests were run on the same MTS Satec frame in axial stress control at a frequency of 5 Hz with R=0.1. Tests were run until failure or runout, which was defined as greater than 2,000,000 cycles. Roughness was measured with a laser confocal microscope.
Rotopolishing and SILC cleaning improved the failure strain (15-21%) of solid and surface porous samples due to the decreased roughness, 0.38 μm and 1.27 μm, respectively. As expected, porous samples had lower properties compared to solid and surface porous samples, regardless of post-processing treatment. Fatigue strength was dependent upon surface roughness, where the smoothest surface (0.38 μm) had the highest fatigue strength (450 MPa). As surface roughness increased (0.38 to 6 μm), the fatigue strength decreased from 450 MPa to 200 MPa due to increased number of sites for crack initiation. Once the porous structure is the dominant structural feature, the impact of post-processing treatment on fatigue is minimal. For SLM devices under high loading conditions, surface porosity can be added without a large decrease in monotonic properties, but fatigue strength will decrease due to the porosity even with a post-processing treatment. These results will be discussed in the context of an additively manufactured, FDA 510(k) cleared, orthopaedic device for foot and ankle applications.