Fracture frequency of all-ceramic crowns during dynamic loading in a chewing simulator using different loading and luting protocols.

OBJECTIVE: The purpose of this laboratory study was to compare the frequency of failures (complete fractures or partial cracks) of molar crowns made of two different all-ceramic materials during dynamic loading in a chewing simulator, as well as the fracture load when subjected to static loading, in relation to different dynamic loading and luting protocols.
METHODS: One hundred and forty-four molar crowns fabricated with IPS Empress or an experimental e.max Press material with high translucency (e.max Press Exp) were luted on CAD/CAM milled PMMA abutments (first lower molar, circular chamfer) either with Variolink or glass-ionomer cement (GIC). All crowns were loaded according to three different loading protocols (n=12 per group) and two force profiles (sinusoidal, rectangular) in a pneumatically driven chewing simulator with a steel stylus (Ø 8 mm) and they underwent simultaneous thermocycling (5 degrees C/55 degrees C). After each phase the crowns were evaluated with regard to fractures or cracks. After dynamic testing, the crowns that did not fail were subjected to compression loading until complete fracture in a universal testing machine (UTM). As control groups, unloaded crowns were also subjected to a UTM. Survival statistics with log-rank tests were applied for the results of the dynamic loading, while ANOVA with post hoc Tukey B was used for the fracture load results and two-way ANOVA was carried out for logarithmically transformed data. Weibull statistics were calculated for pooled fracture load data of the dynamically loaded and control groups.
RESULTS: In the 144 IPS Empress crowns, complete fractures were observed in 9 crowns and partial cracks in another 3 crowns. When the data was pooled, a statistically significant increase in fractures occurred when the sinusoidal force profile was applied compared to a rectangular force profile (log-rank, p<0.05). No fractures occurred in the e.max Press Exp crowns. The two-way ANOVA showed that the type of luting protocol used had the most significant effect on the fracture load of both materials. In conjunction with Empress, however, the luting material influenced the variability twice as much as in e.max Press Exp. There was no statistically significant difference in the fracture load of GIC-luted e.max Press Exp crowns and that of the Variolink luted Empress crowns. The force profile had a significant effect on the fracture load only of the Empress crowns but not of the e.max Press Exp crowns. Weibull statistics revealed a higher scattering of the data of dynamically loaded crowns compared to that of the control groups.
CONCLUSIONS: For testing all-ceramic materials, dynamic loading is indispensable to draw valid conclusions on clinical performance of all-ceramic molar crowns. A sinusoidal profile is advisable, while a gradual increase of the force amplitude does not significantly affect the results.