Objectives: Verify the influence of specimen dimensions on composite shrinkage stress in testing systems of known compliance and in situations where axial strain of bonding substrates was suppressed. Stress distribution was evaluated using finite element analysis (FEA).
Methods: A chemically activated composite (Bisfill 2B, Bisco) was inserted between the flat surfaces of two glass rods (2, 4 or 6mm diameter, D) attached to a universal testing machine. Specimen height (H) was defined by adjusting the distance between the rods (0.5, 1, 2 or 4mm). An extensometer was used to monitor the distance between them. Maximum force after 30min of polymerization was divided by the cross-sectional area of the rod to obtain nominal stress (sigma(n)). Mathematical equations were employed in order to estimate the 'corrected nominal stress' (sigma(cor)) that would be obtained in ideally rigid systems. Data were analyzed by two-way ANOVA/Tukey test (alpha=0.05) and regression analysis (stress versus 'C factor' and stress versus specimen volume). Axysimetrical 2D models were used to evaluate X-component stress distribution (sigma(x)) in the zero compliance condition.
Results: The interactions between D and H were significant for sigma(n) and sigma(cor) (p<0.001). For D=2mm, height influenced only sigma(cor) values. A direct correlation was found between stress and C factor (sigma(n):R(2)=0.959; sigma(cor):R(2)=0.923), but not between stress and volume. FEA evidenced the effect of boundary restraints on sigma(x).
Significance: Specimen dimensions influenced test results, mostly regarding sigma(cor). Stress values strongly increased with the confinement of the specimen probably due to stress concentration adjacent to the bonded interface.