Effects of occlusal cavity configuration on 3D shrinkage vectors in a flowable composite

The objective of this study was to investigate the effects of cavity configuration on the shrinkage vectors of a flowable resin-based composite (RBC) placed in occlusal cavities.

Twenty-seven human molars were divided into three groups (n = 9) according to cavity configuration: "adhesive," "diverging," and "cylindrical." The "adhesive" cavity represented beveled enamel margins and occlusally converging walls, the "diverging" cavity had occlusally diverging walls, and the "cylindrical" cavity had parallel walls (diameter = 6 mm); all cavities were 3 mm deep. Each prepared cavity was treated with a self-etch adhesive (Adper Easy Bond, 3 M ESPE) and filled with a flowable RBC (Tetric EvoFlow, Ivoclar Vivadent) to which had been added 2 wt% traceable glass beads. Two micro-CT scans were performed on each sample (uncured and cured). The scans were then subjected to medical image registration for shrinkage vector calculation. Shrinkage vectors were evaluated three-dimensionally (3D) and in the axial direction.

The "adhesive" group had the greatest mean 3D shrinkage vector lengths and upward movement (31.1 ± 10.9 μm; - 13.7 ± 12.1 μm), followed by the "diverging" (27.4 ± 12.1 μm; - 5.7 ± 17.2 μm) and "cylindrical" groups (23.3 ± 11.1 μm; - 3.7 ± 13.6 μm); all groups differed significantly (p < 0.001 for each comparison, one-way ANOVA, Tamhane's T2).

The values and direction of the shrinkage vectors as well as interfacial debonding varied according to the cavity configuration.

Cavity configuration in terms of wall orientation and beveling of enamel margin influences the shrinkage pattern of composites.