Speaker
Description
Background: Standardized fatigue testing is required for dental implants to estimate the mechanical quality and lifespan of these implants under cyclic loading conditions defined by ISO [1]. However, this testing represents a significant expense in terms of cost and time spent. It is why the present work proposes a predictive method to reduce this cost. This approach is mainly based on the combination of numerical modelling by FEM with mathematical models, such as fatigue life models and fatigue criteria.
After a review on different existing works and models developed for prediction of fatigue behaviour, we finally selected Szajek's model based on Goodman's fatigue theory and Basquin's fatigue life model [2], because this model is simplest and its computing time is reasonable.
To validate this approach, the predictive results are compared to experimental fatigue testing results given by CETIM on LIKE-BS dental implants provided by LIKE Implants Co.
Methods: An FE model is built to get the stress distribution due to the loadings applied to the implant during fatigue tests. Concerning the loading, each step has the maximum and minimum loads according to the cyclic fatigue testing. The screw connecting the prosthetic and the implant body is preloaded by a torque equal to 35 N.cm. This tightening torque is represented by pre-stress according to the screw’s shape, threads and friction coefficients [3]. After a convergence test, the FE model has 195000 tetrahedral or hexahedral elements (4 nodes or 8 nodes per element).
All the contact surfaces between components are considered perfectly linked, except the one between the prosthetic part and the implant body and another between the screw and the prosthetic part.
After analyses with the global FE model, a sub-modelling technic is used to get finer meshes on the crucial part, i.e. the screw, and to reduce the computing efforts.
Szajek’s model is based on modified Goodman’s theory and Basquin’s fatigue life [2].
Finally, it could be enough to obtain principal stresses on critical part of the structure to predict the cycle number of fatigue rupture corresponding to the cyclic loading.
Results: Three different screw friction coefficients are used to compare to the experimental data: 0.6, 0.5 and 0.4 (see Fig. 2). The value of 0.5 gives the best prediction for the lifespan of the dental implant. This value allows defining the preload force due to screw’s tightening torque, equal to 340 N. By using Szajek’s model, the lifespan for each cyclic loading is obtained and compared to experimental data given by CETIM. More detailed results would be shown in oral presentation.
Conclusion: Szajek's model gives good results in comparison to experimental data. It means that we could reduce much time and cost to design new dental implants in the future. Different friction coefficients, which is always ambiguous, was studied. Moreover, the sub-modelling technique can reduce computational costs and give numerical results more accurate.
In near future, it could be interesting to study other models of dental implants to verify whether Szajek's model remains valid for the lifespan prediction.
References
[1] ISO 14801:2016, Médecine bucco-dentaire —Implants — Essai de charge dynamique pour implants dentaires endo-osseux, pp. 20, 2016-11.
[2] K. Szajek, Optimization of a two-component implantology system using genetic algorithm, Ph.D dissertation, Poznan University of Technology, 2013.
[3] V. Stolyarov, "Reduction of friction coefficient of ultrafine-grained CP titanium," Materials Science and Engineering, pp. 313-317, 2004.
[4] D. Flanagan, H. Ilies, P. McCullough, S. McQuoid, Measurement of the fatigue of a mini dental implant: A pilot Study, Journal of Oral Implantology, 34 (1), pp. 7-11, 2008.
| Keywords | Dental implants, Fatigue behaviour, Predictive method, Combined approach |
|---|
