Methods: Fifty sets of two stainless-steel abutments were randomly assigned to five groups (n = 10 each) depending on the material and technique used for manufacturing the FDPs: (1) Metal-ceramic (MC, control); (2) Lava Zirconia (LZ, bi-layered); (3) Lava Plus (LM, monolithic); (4) VITA In-Ceram YZ (YZ, bi-layered); and (5) IPS e-max ZirCAD (ZZ, bi-layered). After being luted to the dies, all FDPs were submitted to thermo-mechanical cycling (120,000 masticatory cycles, 50 N; plus 774 thermal cycles of 5 degrees C/55 degrees C, dwell time: 30 s). Samples were then subjected to a three-point bending test until fracture in a universal testing machine (cross-head speed: 1 mm/min). Fracture load of the veneering ceramic (VF) and total fracture load (TF) were recorded. Microstructure and failure patterns were assessed. Data was analysed using one-way ANOVA and Tukey HSD post-hoc tests (alpha = 0.05).
Results: MC restorations recorded higher VF and TF values than did zirconia FDPs (p = 0.0001), which showed no between-group differences. Within the bi-layered groups, TF was significantly higher than VF. LM pieces registered lower average grain size than did LZ specimens (p = 0.001). Overall, the connector was the weakest part.
Conclusions: All of the groups tested could withstand clinical chewing forces in terms of average fracture load. Zirconia-based samples performed similarly to each other, but showed lower mean fracture load values than did metal-ceramic ones.
Clinical significance: Monolithic zirconia may be recommendable for solving the chipping problem.