The present study aimed to evaluate the impact of the existence of an abutment screw-access hole and the filling effects on the fatigue mechanical behavior of a luted lithium-disilicate glass-ceramic. Seventy-two discs (Ø = 10 mm, 1.0 mm in thickness) of lithium disilicate (IPS e.max CAD, Ivoclar AG) were obtained from prefabricated blocks. Thirty-six abutment specimens of an opaque zirconia (Yz - IPS e.max ZirCAD, Ivoclar AG) and titanium (Ti - Luminesse Ti-Cam discs, Talladium Inc.) were confectioned, and allocated according to 6 groups: Yz and Ti rigid (without screw access hole); Yz unfilled, Yz filled, Ti unfilled and Ti filled (with the screw access (Ø = 2.5 mm) in the center). For the unfilled groups, only a polytetrafluoroethylene tape was used. Resin composite (Tetric N-Ceram, Ivoclar AG) was applied to the screw access hole for the filled groups (Yz and Ti). A cyclic fatigue test was carried out (load of 200 N, 10,000 cycles each; 20 Hz of frequency, step size of 100 N until failure detection (radial/cone crack). The fatigue failure load (FFL) and number of cycles until failure (CFF) were recorded for statistical purposes. The stress distribution (MPa) was evaluated by finite element analysis. A statistically positive effect of the abutment material and the presence of the screw access hole was observed (p ≤ 0.05). The rigid groups (without screw access holes) depicted almost 100% of survival after the fatigue tests. Among the other groups, the Yz-filled group showed the best performance (p ≤ 0.05), followed by the Yz unfilled group. The Ti groups depicted lower values of FFL and CFF, with the Ti unfilled group showing the most unfavorable fatigue behavior (p ≤ 0.05). The lowest tensile stress concentration in the restorative material was observed with the use of rigid abutments, the filled groups depicted intermediate values, while unfilled groups showed the highest stress concentration (Yz rigid = 306.3 MPa; Ti rigid = 310.4 MPa < Yz filled = 490.7 MPa; Ti filled = 498.9 MPa < Yz unfilled = 707.6 MPa; Ti unfilled = 719.7 MPa). Therefore, the presence of a screw-access hole decreases the mechanical performance of a lithium disilicate ceramic regardless of the abutment material. In the presence of a screw-access hole, zirconia abutments depicted a higher fatigue failure load when compared with titanium. The filling of the abutment screw-access hole with resin composite increased the mechanical performance of the simulated restoration.

This study evaluated the effect of substrate core materials and occlusal contact patterns on the fatigue mechanical behavior and stress distribution of single-unit ceramic crowns. One hundred and twenty monolithic crowns were fabricated from zirconia (YZ - IPS e.max ZirCAD, Ivoclar), lithium disilicate (LD - IPS e.max CAD, Ivoclar) and polymer infiltrated ceramic network (PICN - Enamic, Vita Zahnfabrik). The crowns were allocated considering two factors: 'substrate' (epoxy resin or cast Ni-Cr metal core) and 'occlusal contact pattern' (contact at the cusp ridges or cusp tips). The substrate models were design, milled and scanned to plan the restorations in a digital workflow. The crowns were milled, bonded to the substrates, and subjected to an accelerated fatigue test (100 N; 10,000 cycles/step; 20 Hz step-size: 100 N up to 1600 N, and after, 200 N until failure or survival at 2,800N; immersed in water). Statistical analyses were performed using two-way ANOVA, Tukey's post-hoc test, and Kaplan-Meier survival analysis (α = 0.05) considering fatigue failure load and cycles for fatigue failure (FFL/CFF). Fractographic and finite element analysis (FEA) were carried out. The results indicate that the 'substrate' factor did not influence the mechanical behavior of YZ and LD monolithic crowns (p > 0.05). However, PICN crowns bonded to epoxy resin exhibited statistically superior results for FFL and CFF (p < 0.05) compared to Ni-Cr cores. Regarding the 'occlusal contact pattern' factor, YZ and LD exhibited higher mean FFL and CFF when associated with cusp tip contact compared to cusp ridge contact (p < 0.05), except for YZ bonded to the epoxy resin substrate (p > 0.05). No differences were detected for the 'occlusal contact' factor in PICN crowns (p > 0.05). The predominant failure was Hertzian cone cracks, regardless of the restorative material. Stress measurements showed higher stress peaks at the cusp ridges. The core material did not alter the fatigue mechanical behavior of YZ or LD crowns. However, the incidence of cusp ridge contacts in YZ or LD crown increases the risk of failure. Conversely, when using PICN crowns, a core with a more similar elastic modulus enhances mechanical behavior compared to a stiffer core, and no influence on the occlusal pattern was observed.