The equivalent stress values of the contact model were higher at the implant neck than the tip, and the stress generation range was also broader. However, in the fixation model, the implant neck and tip had substantially equivalent values and the stress generation range was also narrower than that of the contact model. This shows that under immediate loading conditions, there is a high likelihood that loading applied to the superstructure is also supported by cancellous bone at the implant threads and tip, but the majority is supported by cortical bone at the implant neck. That is, under immediate loading conditions, it is believed there is a need to be mindful of the stress concentration at the implant neck. The material properties of the FEA models constructed in the present study cannot not be compared to actual oral cavity stress values because they differ significantly from an actual bone. However, when regarding how peri-implant bone is impacted by contact between the implant and bone, it is considered sufficiently useful in predicting tendencies.

The CV of the equivalent stress values was calculated to assess the reproducibility of the contact and fixation models. The CV ranged from 0.52 to 45.99, showing a tendency for higher overall values compared with the CV for displacement under loading. The reproducibility of the equivalent stress values had considerable variance from model to model in some regions. In particular, the contact model showed a higher CV at both the neck and tip. In an analysis of contact conditions, moving the nodes also dramatically changed the stress and strain occurring at the interface; a stress concentration was also generated depending on the shape of the model [40]. Though the FEA models were fabricated under similar conditions, it appears that a minute error in shape caused in the element divisions appeared in the form of a large error in equivalent stress values. It is necessary to verify whether the numerical stress values obtained from the FEA have validity by attaching a strain gauge to an experimental model and observing correlations in stress values between the experimental and FEA models.