3D statistical failure analysis of monolithic dental ceramic crowns

Computation of the fracture probability and lifetime of all ceramic anterior crowns under cyclic loading - An FEA study

OBJECTIVES

To predict the lifetime and fracture probability of anterior crowns made of a lithium disilicate glass-ceramic (IPS e.max CAD, LD, Ivoclar Vivadent, Liechtenstein) and a zirconia-containing lithium silicate glass-ceramic (Celtra Duo, ZLS, Dentsply Sirona, USA) under cycling loading.

METHODS

Three-point bending tests were conducted to measure the viscoelastic parameters. These parameters are used to compute the residual stresses of the anterior crown after crystallization. In the next analysis, the cyclic loading on the anterior crown was calculated. Based on this combined stress state (residual stress and stress state due to external cyclic loading), the life cycle and fracture probability of the anterior crown was calculated using the CARES/Life software. Finally, fatigue experiments were carried out to compare and validate the results of the computations.

RESULTS

Although a sound qualitative comparison of the lifetime of both materials can be done using this methodology, the calculated fracture probability of the anterior crown for both materials was very low in comparison with the fatigue test results using the fatigue parameters determined from the experiments. In order to achieve good correspondence with the experimental results, the SCG exponent n for both materials should be modified by a correlation factor of 0.38.

SIGNIFICANCE

Using this modified computational strategy, the results of the time-consuming fatigue tests for dental glass-ceramics can be closely predicted. This methodology can be integrated into the development process of new glass-ceramic materials in order to save time and costs.

The marginal gaps of lithium disilicate crowns constructed by different scanner and milling unit combinations

BACKGROUND

The purpose of this study was to compare the marginal gaps of CAD/CAM lithium disilicate (LDS) crowns constructed using three different scanner and milling unit combinations.

METHODS

Twenty-four Columbia model lower left molars were prepared for LDS crowns by undergraduate students in a simulated environment. One LDS crown was constructed for each crown preparation using each of the following intraoral scanner/milling unit combinations: TRIOS 3 scanner/E4D milling unit (TRIO/E4DM); TRIOS 3 scanner/Sirona inLab MC X5 milling unit (TRIO/MCX5); E4D scanner/E4D milling unit (E4DS/E4DM). The crowns were seated onto the original crown preparations and three vertical marginal gap measurements taken at four locations (mid-buccal, mid-lingual, mid-mesial and mid-distal) using a stereomicroscope. The mean marginal gap (MMG) was calculated for each crown and each individual tooth surface.

RESULTS

The TRIO/MCX5 scanner/milling unit combination produced the smallest MMG of 63.73 ± 47.38 μm followed by E4DS/E4DM (88.64 ± 106.51 μm) and TRIO/E4DM (107.41 ± 76.47 μm). There was a significant effect of milling unit (p < 0.0001) on MMG but no significant effect of scanner (p = 0.070) or location (p = 0.182).

CONCLUSIONS

The newer scanner/milling unit combination produced LDS crowns with the smallest MMG. All scanner/milling unit combinations produced LDS crowns within a clinically acceptable MMG.

24 hour polymerization shrinkage of resin composite core materials

PURPOSE

The study's purpose was to evaluate the 24-hour polymerization shrinkage of resin composite core materials.

MATERIAL AND METHODS

Eleven resin composite core material samples (n = 12) were evaluated using a non-contact imaging device with measurements obtained over 24 h. Shrinkage values were determined corresponding to proposed times involved with CAD/CAM same-day treatment and at 24 h. Shrinkage data was statistically compared using Friedman/Dunn's test for intragroup analysis and Kruskal Wallis/Dunn's test for intergroup analysis, all at a 95% level of confidence (α = 0.05).

RESULTS

Mean results identified a wide volumetric shrinkage range with considerable similarity overlap. Inconsistent shrinkage behavior was evident and all materials reached maximum values before 24 h. No significant difference was observed during proposed digital same day all ceramic crown procedures, but some differences were noted at 24 h.

CONCLUSIONS

Under this study's conditions results were material specific, at times inconsistent, with wide variation. Shrinkage consistently increased for all products and it is not known if the continued shrinkage magnitude may compromise the stability and fit of all-ceramic crowns at 24 h.

Biomaterials: Ceramic and Adhesive Technologies

This review highlights ceramic material options and their use. The newer high-strength ceramics in monolithic form have gained popularity despite the lack of long-term clinical data to support this paradigm shift. Although there are some encouraging clinical data available, there is a need to develop laboratory simulation models that can help predict long-term clinical performance for ceramic and adhesive cements.

The effect of adhesive failure and defects on the stress distribution in all-ceramic crowns

OBJECTIVES

To explore the effect of adhesive failure and defects between the crown and cement on the stress distribution within all-ceramic crowns and the corresponding risk of failure.

METHODS

An IPS e.max crown of lithium disilicate produced by CAD/CAM for a first mandibular molar was modeled using finite element analysis based on X-ray micro-CT scanned images. Predefined debonding states and interfacial defects between the crown and cement were simulated using the model. The first principal stress distribution of the crown and cement was analyzed under a vertical occlusal load of 600 N. A concept of failure risk was proposed to evaluate the crown.

RESULTS

Stress concentrations in the crown were identified on the occlusal surface surrounding the region of loading, beneath the area of loading and at the margin of the interior surface. Stress concentrations in the cement were also evident at the boundary of the debonded areas. The lower surface of the crown is safe to sustain the 600 N vertical load, but the top surface of the cement would undergo cohesive failure. According to the evaluation of failure risk of the crown, the conditions of highest risk corresponded to the conditions with highest percentage of cement damage. The risk of failure is not only associated with debonding between the crown and cement, but also associated with its distribution.

CONCLUSIONS

Debonding related defects and cementing defects are more deleterious to the interfacial stress than debonding itself. The axial wall plays a critical role in maintaining the principal tensile stress of the crown at an acceptable level.

Approximate relative fatigue life estimation methods for thin-walled monolithic ceramic crowns

OBJECTIVES

The objective is to establish an approximate relative fatigue life estimation method under simulated mastication load for thin-walled monolithic restorations.

METHODS

Experimentally measured fatigue parameters of fluormica, leucite, lithium disilicate and yttrium-stabilized zirconia in the existing literature were expressed in terms of the maximum cyclic stress and stress corresponding to initial crack size prior to N number of loading cycles to assess their differences. Assuming that failures mostly originate from high stress region, an approximate restoration life method was explored by ignoring the multi-axial nature of stress state. Experiments utilizing a simple trilayer restoration model with ceramic LD were performed to test the model validity.

RESULTS

Ceramic fatigue was found to be similar for clinically relevant loading range and mastication frequency, resulting in the development of an approximate fatigue equation that is universally applicable to a wide range of dental ceramic materials. The equation was incorporated into the approximate restoration life estimation, leading to a simple expression in terms of fast fracture parameters, high stress area ΔA, the high stress averaged over ΔA and N. The developed method was preliminarily verified by the experiments. The impact of fast fracture parameters on the restoration life was separated from other factors, and the importance of surface preparation was manifested in the simplified equation. Both the maximum stress and the area of high stress region were also shown to play critical roles.

SIGNIFICANCE

While nothing can replace actual clinical studies, this method could provide a reasonable preliminary estimation of relative restoration life.

Fatigue resistance of all-ceramic fixed partial dentures - Fatigue tests and finite element analysis

OBJECTIVE

To estimate the fatigue resistance of a new translucent zirconia material in comparison to lithium disilicate for 3-unit fixed partial dentures (FPDs).

METHODS

Eighteen 3-unit FPDs (replacement of first upper molar) with a connector size of 4mm×4mm were dry milled with a five-axis milling machine (Zenotec Select, Wieland, Germany) using discs made of a new translucent zirconia material (IPS e.max ZirCAD MT, Ivoclar Vivadent). Another 9 FPDs with a reduced connector size (3mm×4mm) were milled. The zirconia FPDs were sintered at 1500°C. For a comparison, 9 FPDs were made of IPS e.max Press, using the same dimensions. These IPS e.max Press FPDs were ground from a wax disc (Wieland), invested and pressed at 920°C. All FPDs were glazed twice. The FPDs were adhesively luted to PMMA dies with Multilink Automix. Dynamic cyclic loading was carried out on the molar pontic using Dyna-Mess testing machines (Stolberg, Germany) with 2×10⁶ cycles at 2Hz in water (37°C). Two specimens per group and load were subjected to decreasing load levels (at least 4) until the two specimens no longer showed any failures. Another third specimen was subjected to this load to confirm the result. All the specimens were evaluated under a stereo microscope (20× magnification). The number of cycles reached before observing a failure, and their dependence on the load and on the material, were modeled, using a Weibull model. This made it possible to estimate the fatigue resistance as the maximum load for which one would observe less than 1% failure after 2×10⁶ cycles. In addition to the experimental study, Finite Element Modeling (FEM) simulations were conducted to predict the force to failure for IPS e.max ZirCAD MT and IPS e.max Press with a reduced cross-section of the connectors.

RESULTS

The failure mode of the zirconia FPDs was mostly the fracture of the distal connector, whereas the failure mode of the lithium disilicate FPDs observed to be the fracture of the connectors or multiple cracks of the pontic. The fatigue resistance with 1% fracture probability was estimated to be 488N for the IPS e.max ZirCAD MT FPDs (453N for repeated test), 365N for IPS e.max ZirCAD MT FPDs with reduced connector size and 286N for the e.max Press FPDs. All three IPS e.max ZirCAD groups statistically performed significantly better than IPS e.max Press (p<0.001). On the other hand, no significant difference could be established between the two IPS e.max ZirCAD MT3 groups with a 4mm×4mm connector size (p>0.05). The allowable maximum principal stress (σ_max) which did not lead to failure during fatigue testing for IPS e.max ZirCAD MT3 was calculated between 208MPa and 223MPa for FPDs with 4mm×4mm connectors for 2×10⁶ cycles. This value could also be verified for the FPDs of the same material with 3mm×4mm connectors. On the other hand fatigue strength in terms of σ_max at 2×10⁶ cycles of IPS e.max Press was calculated to be between 78 and 90MPa.

SIGNIFICANCE

The fatigue resistance of the translucent zirconia 3-unit FPDs was about 60-70% higher than that of the lithium disilicate 3-unit FPDs, which may justify their use for molar replacements, provided that a minimal connector size of 4mm×4mm is observed. Even with a limited number of specimens (n=9) per group it was possible to statistically differentiate between the tested groups.

Survival Predictions of Ceramic Crowns Using Statistical Fracture Mechanics

This work establishes a survival probability methodology for interface-initiated fatigue failures of monolithic ceramic crowns under simulated masticatory loading. A complete 3-dimensional (3D) finite element analysis model of a minimally reduced molar crown was developed using commercially available hardware and software. Estimates of material surface flaw distributions and fatigue parameters for 3 reinforced glass-ceramics (fluormica [FM], leucite [LR], and lithium disilicate [LD]) and a dense sintered yttrium-stabilized zirconia (YZ) were obtained from the literature and incorporated into the model. Utilizing the proposed fracture mechanics-based model, crown survival probability as a function of loading cycles was obtained from simulations performed on the 4 ceramic materials utilizing identical crown geometries and loading conditions. The weaker ceramic materials (FM and LR) resulted in lower survival rates than the more recently developed higher-strength ceramic materials (LD and YZ). The simulated 10-y survival rate of crowns fabricated from YZ was only slightly better than those fabricated from LD. In addition, 2 of the model crown systems (FM and LD) were expanded to determine regional-dependent failure probabilities. This analysis predicted that the LD-based crowns were more likely to fail from fractures initiating from margin areas, whereas the FM-based crowns showed a slightly higher probability of failure from fractures initiating from the occlusal table below the contact areas. These 2 predicted fracture initiation locations have some agreement with reported fractographic analyses of failed crowns. In this model, we considered the maximum tensile stress tangential to the interfacial surface, as opposed to the more universally reported maximum principal stress, because it more directly impacts crack propagation. While the accuracy of these predictions needs to be experimentally verified, the model can provide a fundamental understanding of the importance that pre-existing flaws at the intaglio surface have on fatigue failures.