The use of Weibull statistics in mechanical testing of brittle dental materials

Full mapping tensile bond strength of luting in search for differences due to centripetal curing shrinkage

OBJECTIVES

testing if hypothetical transverse centripetal strains due to polymerization contraction of luting materials produce differential alterations in its bonding to luted structures, depending on distances to the center of the luting mass, and if this effect is C-factor related. Two hypotheses were tested: (1) there is a statistically significant decreasing relationship between the bonding strength and the transverse distances to the center of the luting material, and (2) there is a statistically significant difference between bonding strengths among luting spaces with different configuration factors.

METHODS

10 PMMA (15 mm Ø) pairs of cylinders were cemented (Scotchbond Universal adhesive & Relyx Universal, both chemically cured) in a compliant setup under two (20 and 70 N) luting forces forming 2 groups (5 samples each), resulting in different C-factors. Whole samples were sectioned in x and y directions obtaining non-trimmed beams from all along the luting surfaces. Their relative positions in each sample were assessed before separating and categorized (10 categories) according to their distances to the center of the sample. All beams were tested in tension and, because of their uneven bonding areas and to balance its influence, UTS results were transformed into UTS_res. First hypothesis was tested trough a linear relationship between UTS_res and distances to vertical centers per samples. Second hypothesis was tested using Mann-Whitney U tests to compare UTS_res between groups, along all categories. Further Weibull analysis was applied.

RESULTS

ANOVA's p of the regression UTS_res - categories were statistically significant for all samples in group 70 N and for all except one in group 20 N: first hypothesis is partially maintained. Although Mann-Whitney tests p comparing UTS_res of both groups for all categories but the first were statistically significant this hypothesis was maintained relying in Weibull analysis.

SIGNIFICANCE

bonded attachment of cemented materials decreases from centers to outbounds in plane, extensive surfaces, and this decrease is C-factor related.

Load bearing capacity and Weibull characteristics of inlay-retained resin-bonded fixed dental prosthesis made of all-ceramic, fiber-reinforced composite and metal-ceramic after cyclic loading

PURPOSE

The objectives of this study were to evaluate the load bearing capacity of RBFDPs made of different materials after cyclic loading, and classify the failure types after loading.

MATERIALS AND METHODS

Sound human mandibular first premolars and first molar pairs (N = 60 per tooth type, n = 10 per group) were randomly divided into six experimental groups to receive one of the following inlay-retained RBFDP types: RC: Resin composite only, DFRC: Direct E-glass fiber-reinforced composite (FRC), IFRC: Indirect E-glass FRC, LS: Lithium disilicate glass-ceramic, ZR: Yttria-stabilized tetragonal zirconia, MC: Metal-ceramic. Box preparations were made in abutment teeth using diamond burs followed by standardized ultrasonic burs. The teeth were conditioned employing an etch-and-rinse adhesive system and the indirect RBFDPs were cemented adhesively. The specimens were subjected to cyclic loading for x1.200.000 in distilled water alternating between 5 and 55 °C (Zurich Chewing Simulator). They were then loaded to failure from the occlusal surface in the Universal Testing Machine (cross-head speed: 1 mm/min). Failure types were classified as repairable or irreparable depending on the location and size. Data were analyzed using Welch and Tamhane's T2 post-hoc tests (α = 0.05). Weibull modulus for each group was calculated based on parametric distribution analysis of censored data for maximum fracture load.

RESULTS

Mean load bearing capacity (N) of Groups LS (1274 ± 270), ZR (1567 ± 363) and MC (1544 ± 787) were significantly higher than those of other groups (p < 0.05). RC, DFRC, IFRC (601 ± 130 - 819 ± 270) and MC did not show significant difference (p > 0.05). Weibull modulus (m) was the highest in Group LS (m = 5.3) followed by Group RC (m = 5.1). Other groups presented Weibull moduli ranging between 1.4 and 3.3. Only in Group ZR, 2 early debonding occurred during cyclic loading. While in this group predominantly irreparable failures (debonding with without tooth fracture) were observed (8 out of 10), all other groups presented mainly single or a combination of repairable failures (chipping in the veneering material). DFRC, IDRC and MC did not show any debonding from the abutment teeth.

CONCLUSION

Considering load bearing capacity, repairable failure types and Weibull moduli, lithium disilicate seem to be more durable than those of other material options for posterior inlay-retained RBFDPs. Due to early debondings and catastrophic irreparable failure types, zirconia RBFDPs should be indicated with caution in the posterior region.

The influence of clinically induced variability on the bi-axial fracture strength of cemented aluminous core porcelain discs

OBJECTIVES

Zinc phosphate cements are supplied in powder/liquid form so that a range of mixing ratios are naturally generated in clinical practice. However, these cements are known to have a deleterious effect on dental ceramics and the impact of variations in cement mixing on the strength of dental porcelain was investigated in the present study.

METHODS

Vitadur-N core porcelain was fabricated into discs (15 mm diameter and 2 mm thickness), zinc phosphate cements of different mixing ratios were applied to the discs to produce a luting cement thickness and the samples were stored for 24 h prior to testing in a water-bath at 37 +/- 1 degrees C. The mean fracture strength, standard deviation and associated Weibull moduli (m) of the coated porcelain discs were determined as a function of mixing ratio using bi-axial flexure (ball-on-ring) by fracturing sets of 30 specimens coated with cements manipulated at different mixing ratios.

RESULTS

The strength data for the coated porcelain discs showed little variation in magnitude and consistency ranging from 108 +/- 8.3 MPa (m = 14.0 +/- 2.6) at 1.7 g/ml to 112 +/- 8.5 MPa (m = 14.6 +/- 2.7) at 2.6 g/ml and 108 +/- 8.1 MPa (m = 14.4 +/- 2.6) at 3.2 g/ml. However, the plots of survival probability against strength for coated specimens prepared using mixing ratios below that recommended by the manufacturers' for luting purposes (2.6 g/ml), appeared to develop an asymmetry at the lower values of strength.

SIGNIFICANCE

It was proposed that the increased corrosive acidic environment for cements prepared at mixing ratios below 2.6 g/ml may have extended pre-existing flaws in the porcelain discs thereby possibly producing the asymmetry in the survival distributions.