The measured hardness profiles of the implanted and unimplanted specimens are shown in Fig. 3. The data for the unimplanted specimen are grouped fairly tightly (exhibiting relatively small experimental error) and they show a slight rise as the depth of the indent approaches zero. The hardness data for the implanted specimen shows greater error than the unimplanted case, especially at the shallow depths, but the hardness clearly exceeds that of the unimplanted case. At a depth of 25 nm, the hardness of the implanted specimen is 3-5 times that of the unimplanted specimen. The peak relative hardness, which occurs right at the surface, could not be measured because the error in the measurements was unreasonable at depths shallower than 25 nm. The curves in Fig. 3 are the results of the finite element simulation. The solid line gives the results for the unimplanted specimen. It shows reasonable agreement, although it exhibits a slight decrease with decreasing depth, rather than the slight increase exhibited by the data. The dotted line in this figure shows the results of finite element simulations for implanted specimens. This particular result is for a case in which the yield stress at the surface of the model was 10 times that of the bulk yield stress. The agreement is reasonable, indicating that the nitrogen ion implantation yielded titanium nitrides (or some other phases) which have yield stresses about 10 times that of the bulk Ti-6Al-4V alloy. (For comparison increases of a factor of 40 have been seen in ion implantation of aluminum [8]). Assuming that the bulk yield stress is approximately 825 MPa, this indicates that the yield stress of the implanted layer is on the order of 5-8 GPa. Efforts are under way to estimate the strength of titanium nitride by some other means, in order to provide some confirmation of this result and to verify the assumption that there is in fact a significant amount of titanium nitride in the implanted layer.
In the simulation results, the profile of the hardness is determined by the profile of the yield stress. Since the hardness profiles from the simulation agree with those from the experiment, the combination of the two provides some indication that the assumed yield stress profile is reasonable. Hence, the property changes can be assumed to follow the nitrogen concentration in nitrogen-implanted Ti-6Al-4V. More work is needed to establish the mechanisms by which this profile is achieved.
In order to provide additional validity to the comparison of the finite element simulation to the hardness tests, a sensitivity study was carried out to assess the range of the peak yield stress that ion implantation produced in the Ti-6Al-4V. The peak yield stress in the model was varied, while the shape of the profile was maintained. The results of this study are shown in Fig. 4. The results are given for peak yield stresses which are 1, 4, 6, 8, and 10 times the bulk yield stress. The hardness is shown to increase steadily as the peak yield stress increases. The curve with a peak yield stress of 10 times that of the bulk was chosen as the one providing the best fit to the data from hardness measurements, so it was the curve used in Fig. 3.
One additional result was noticed during the course of this study. The finite element simulations indicated a fundamental difference in the shape of the indent at full load for the implanted and unimplanted specimens, as shown in Fig. 5. This figure shows the shape of the surface of the indentation for three different simulations. The origin of the abscissa represents the centerline of the indenter (assumed to be conical). The indentation of an unimplanted specimen is shown to produce a berm or ridge at the edge of the contact area, while neither of the simulations of implanted specimens show this behavior. Presumably, the increased surface strength in the case of the implanted specimen enhances the importance of deformations in the bulk. This likely increases the elastic deformation in the bulk, providing space below the indent for the material displaced by the indenter. Again, efforts are under way to measure this effect experimentally.