Clinically relevant fracture testing of all-ceramic crowns

Abstract

Objectives

Fracture strength measured in vitro indicates that most all-ceramic crowns should be able to withstand mastication forces. Nevertheless, fractures are one of the major clinical problems with all-ceramic restorations. Furthermore, the fracture mode of all-ceramic crowns observed in clinical use differs from that found in conventional fracture strength tests. The aim of the present study was to develop and investigate a method that simulates clinical fracture behavior in vitro .

Methods

30 crowns with alumina cores were made to fit a cylindrical model with a molar-like preparation design. These crowns were randomly allocated to 3 tests groups ( n = 10). The crowns in group 1 were cemented to abutment models of epoxy and subsequently fractured by occlusal loading without contact damage. The crowns in group 2 were fractured by cementation with expanding cement. The crowns in group 3 were cemented on an abutment model of epoxy split almost in two and fractured by increasing the diameter of the model in the bucco-lingual direction. The fractured crowns were analyzed by fractographic methods and compared to a reference group of 10 crowns fractured in clinical use.

Results

The fracture modes of all the in vitro crowns were similar to clinical fracture modes. The fracture modes in group 1 were most closely matched to the clinical fractures. These crowns also fractured at clinically relevant loads.

Conclusion

Laboratory tests that induce a distortion of the abutment model during occlusal loading without occlusal contact damage can simulate clinical fractures of all-ceramic crowns.

Introduction

Fractographic analyses of all-ceramic crowns that have failed in clinical use reveal fracture features that differ from those usually observed in laboratory studies . Several types of loading devices have been used to assess fracture strength of all-ceramic crowns. In vitro tests usually induce contact damage at the loading point, occlusally for molars and premolars or palatinally for incisors . Clinical crowns on the other hand, rarely reveal any such damage, but rather fracture by cracks initiating of the cervical margin in the core material .

In vitro studies often present fracture strength values that exceed the loads usually present during mastication, indicating that they should not fracture during normal service . Clinical data, on the other hand, reveal that fractures of all-ceramic restorations are prevalent . Thus, it is indicated that the commonly used laboratory tests do not adequately mimic the clinical fracture resistance and fracture features of dental ceramic restorations . In order to understand the behavior of the ceramic materials in vivo , new methods which simulate clinical fractures in vitro , must be established.

The aim of the present study was to develop in vitro test methods that more closely simulate clinically observed fracture behavior compared with the commonly used axial compression test set-up.

Materials and methods

Thirty-one all-ceramic crowns with 0.6 mm thick alumina cores (Procera Crown Alumina, Nobel Biocare, Stockholm, Sweden) were produced to a model of a molar crown preparation with a diameter of 12 mm. A shallow chamfer finish line marked the crown margin. The angle between axial and occlusal surfaces was rounded and the axial surface was tapered at 5° to the central axis. The finish line was given a curvature similar to a molar where the finish line lies more apically on the buccal and oral surfaces than in the approximal areas. The buccal flange extended 1 mm lower than the lingual flange. The copings were veneered with one layer of veneering ceramic (Ducera Allceram, Degudent, Hanau-Wolfgang, Germany). The veneering process followed the recommendation from the manufacturers. One layer of liner material was applied evenly on the outer surface before firing. The crowns were veneered with about 0.7 mm layer occlusally and a gradual decrease in thickness toward the cervical margin. Slight adjustment to the veneering material was performed with a stone if necessary before glazing to achieve a smooth and even layer of veneering porcelain. The crowns had a slight concave occlusal surface without fissures or grooves, and thus similar to a molar tooth with low cusps.

The crowns were inspected in a light microscope at 16 times magnification for marginal fit, margin flaws, cracks or internal contamination of veneering material before allocation to three test groups ( Fig. 1 ). One crown, with a crack at the crown margin visualized by trans-illumination, was excluded from the study. 16 crowns needed careful removal of excess veneering ceramic at the crown margin or just inside the coping to achieve acceptable fit.

Fig. 1
Illustration of the three test groups. (1) Ten crowns were cemented on an epoxy model and softly loaded on the occlusal surface until fracture. (2) Ten crowns were cemented on a brass model made with a cement gap of 90–100 μm. The cement used had an expansion force of up to 78 MPa. Expansion in μm and time to fracture was recorded. (3) Ten crowns were cemented on a split epoxy model. These were softly loaded on the occlusal surface until fracture. The load made the epoxy model separate in bucco-lingual direction due to the conical rod inserted into the center of the hollow model.

The axial compression (group 1): Ten epoxy (Epofix, Struers, Ballerup, Denmark) models identical to the preparation model were produced. On these models 10 crowns were cemented with zinc phosphate cement (De Trey Zink, Dentsply De Trey GmbH, Konstanz, Germany). The crowns were stored in distilled water at 37 °C for 24 (±1) h. All crowns were then loaded occlusally with a stainless steel ball of 30 mm in diameter in a universal test machine (Lloyd Instruments Ltd., Leicester, UK). Loading was performed while the crowns were immersed in water at 37 °C. In order to avoid contact damage at the loading point, the occusal surface and the steel ball were separated by a 3 mm thick ethylene propylene diene rubber disk with a Shore A hardness of 90 (EPDM 90). Load at fracture was recorded.

Expanding cement (group 2): A brass (Hera trainingmetall, Heraus Kulzer GmbH, Hanau, Germany) copy of the abutment was produced to allow a cement gap of 90–110 μm. Ten crowns were subsequently cemented to the model with expanding cement intended for landscaping purposes (Heydi Trollkraft, Frogner, Norway). The cement develops high expansion forces during setting. Diametral expansion of the crown margins in the bucco-lingual direction was recorded by a dilatometer (Tesaotronic, Tesa Technology, Switzerland) connected to a recorder that logged the expansion over time. Time to fracture was registered.

Expanding abutment (group 3): An epoxy abutment model was hollowed with a cone with a cylindrical base from the bottom. The taper was 5°. A conical rod was made to fit the central cavity. The model was then split into two in the mesio-distal direction from the bottom and up to 2 mm from the occlusal surface to create expansion of the model in the bucco-lingual direction when loaded axially. Ten crowns were cemented on this model with zinc phosphate cement and stored in distilled water at 37 °C for 24 h before loading. The crowns were loaded occlusally as described above. Load at fracture was recorded.

Video recordings were made of each test to record the crack propagation during fracture and to detect early crack growth and cracks in the veneering material before total fracture. The video recordings were used to measure the distance the conial rod was slid inside the abutment for group 3. Measurements were taken in duplicate. By simple geometry, the lateral (transverse) expansion at the cervical margin was calculated. The mean of two separate measurements was used in the calculations.

Clinical fractures (reference group): Ten molar crowns with alumina copings that fractured in clinical use, and obtained from dentist in Norway, were used as reference. These have previously been analyzed using the same fractographic methods and the detailed results of these analyses are presented in a previous study .

Fractographic evaluation

All crowns were analyzed using standard fractographic methods and the crack paths and fracture features were registered (ASTM standard) . The fractographic findings were compared to the video recordings to confirm the crack propagation.

Materials and methods

Thirty-one all-ceramic crowns with 0.6 mm thick alumina cores (Procera Crown Alumina, Nobel Biocare, Stockholm, Sweden) were produced to a model of a molar crown preparation with a diameter of 12 mm. A shallow chamfer finish line marked the crown margin. The angle between axial and occlusal surfaces was rounded and the axial surface was tapered at 5° to the central axis. The finish line was given a curvature similar to a molar where the finish line lies more apically on the buccal and oral surfaces than in the approximal areas. The buccal flange extended 1 mm lower than the lingual flange. The copings were veneered with one layer of veneering ceramic (Ducera Allceram, Degudent, Hanau-Wolfgang, Germany). The veneering process followed the recommendation from the manufacturers. One layer of liner material was applied evenly on the outer surface before firing. The crowns were veneered with about 0.7 mm layer occlusally and a gradual decrease in thickness toward the cervical margin. Slight adjustment to the veneering material was performed with a stone if necessary before glazing to achieve a smooth and even layer of veneering porcelain. The crowns had a slight concave occlusal surface without fissures or grooves, and thus similar to a molar tooth with low cusps.

The crowns were inspected in a light microscope at 16 times magnification for marginal fit, margin flaws, cracks or internal contamination of veneering material before allocation to three test groups ( Fig. 1 ). One crown, with a crack at the crown margin visualized by trans-illumination, was excluded from the study. 16 crowns needed careful removal of excess veneering ceramic at the crown margin or just inside the coping to achieve acceptable fit.

Fig. 1
Illustration of the three test groups. (1) Ten crowns were cemented on an epoxy model and softly loaded on the occlusal surface until fracture. (2) Ten crowns were cemented on a brass model made with a cement gap of 90–100 μm. The cement used had an expansion force of up to 78 MPa. Expansion in μm and time to fracture was recorded. (3) Ten crowns were cemented on a split epoxy model. These were softly loaded on the occlusal surface until fracture. The load made the epoxy model separate in bucco-lingual direction due to the conical rod inserted into the center of the hollow model.

The axial compression (group 1): Ten epoxy (Epofix, Struers, Ballerup, Denmark) models identical to the preparation model were produced. On these models 10 crowns were cemented with zinc phosphate cement (De Trey Zink, Dentsply De Trey GmbH, Konstanz, Germany). The crowns were stored in distilled water at 37 °C for 24 (±1) h. All crowns were then loaded occlusally with a stainless steel ball of 30 mm in diameter in a universal test machine (Lloyd Instruments Ltd., Leicester, UK). Loading was performed while the crowns were immersed in water at 37 °C. In order to avoid contact damage at the loading point, the occusal surface and the steel ball were separated by a 3 mm thick ethylene propylene diene rubber disk with a Shore A hardness of 90 (EPDM 90). Load at fracture was recorded.

Expanding cement (group 2): A brass (Hera trainingmetall, Heraus Kulzer GmbH, Hanau, Germany) copy of the abutment was produced to allow a cement gap of 90–110 μm. Ten crowns were subsequently cemented to the model with expanding cement intended for landscaping purposes (Heydi Trollkraft, Frogner, Norway). The cement develops high expansion forces during setting. Diametral expansion of the crown margins in the bucco-lingual direction was recorded by a dilatometer (Tesaotronic, Tesa Technology, Switzerland) connected to a recorder that logged the expansion over time. Time to fracture was registered.

Expanding abutment (group 3): An epoxy abutment model was hollowed with a cone with a cylindrical base from the bottom. The taper was 5°. A conical rod was made to fit the central cavity. The model was then split into two in the mesio-distal direction from the bottom and up to 2 mm from the occlusal surface to create expansion of the model in the bucco-lingual direction when loaded axially. Ten crowns were cemented on this model with zinc phosphate cement and stored in distilled water at 37 °C for 24 h before loading. The crowns were loaded occlusally as described above. Load at fracture was recorded.

Video recordings were made of each test to record the crack propagation during fracture and to detect early crack growth and cracks in the veneering material before total fracture. The video recordings were used to measure the distance the conial rod was slid inside the abutment for group 3. Measurements were taken in duplicate. By simple geometry, the lateral (transverse) expansion at the cervical margin was calculated. The mean of two separate measurements was used in the calculations.

Clinical fractures (reference group): Ten molar crowns with alumina copings that fractured in clinical use, and obtained from dentist in Norway, were used as reference. These have previously been analyzed using the same fractographic methods and the detailed results of these analyses are presented in a previous study .

Fractographic evaluation

All crowns were analyzed using standard fractographic methods and the crack paths and fracture features were registered (ASTM standard) . The fractographic findings were compared to the video recordings to confirm the crack propagation.

Results

Loads at fracture and standard deviations are listed in Table 1 , together with dimensional change for group 2 and 3, and the mode of fracture for all crowns. The macroscopic fracture modes were separated into 3 categories as shown in Fig. 2 ; (A) One main fracture line from one approximal region to the other separating the crown in two pieces along the occlusal groove, (B) A semi-lunar fracture line not including the occlusal surface and (C) An arrested fracture line. The fractographic analyses revealed that all the laboratory crowns except three crowns in group 2 had fractures that started in the cervical margin in the approximal curvature ( Fig. 3 ). All the fractures in the crowns from the clinical reference group started in the cervical margin and among these, seven started in the approximal area. The fractures that were most similar to clinical fractures were found in group 1 ( Fig. 4 ). In group 1, all fractures started in the cervical margin on one of the approximal surfaces, close to the highest point in the curve of the finish line. The crack propagated to the occlusal surface where it often stopped in a slight compression curve or in a bifurcation in the opposite side of the occlusal surface. One of the fractures continued to the other approximal surface with only a slight compression curve and no bifurcations. Secondary cracks in the veneering ceramic were observed in four of the ten crowns in this group.

Table 1
Results of the tests.
Fracture mode Localization of start Start point Axial load at fracture (in N , mean ± sd) Cervical expansion Bucco-Lingual (in μm; mean ± sd)
Axial compression A 5 Approximal ( n = 10) Cervical ( n = 10) 888.2 (151)
B
C 5
Expanding cement A Approximal ( n = 7) Cervical ( n = 9) 12.6 (5.3)
B 10 Buccal ( n = 2) Occlusal ( n = 1)
C Lingual ( n = 1)
Expanding abutment A 4 Approximal ( n = 10) Cervical ( n = 10) 60.1 (18) 12.8 (2.6) (based on calculation)
B
C 6
Clinical references A 8 Approximal ( n = 8) Cervical ( n = 10)
B 2 Buccal ( n = 1)
C** Lingual ( n = 1)
Only gold members can continue reading. Log In or Register to continue

Nov 25, 2017 | Posted by in Dental Materials | Comments Off on Clinically relevant fracture testing of all-ceramic crowns
Premium Wordpress Themes by UFO Themes