| Sign In to gain access to subscriptions and/or personal tools. |
Effect of Surface Roughness on Flexural Strength of Veneer Ceramics
1 Department of Dental Prosthetics, Section of Dental Materials, University RWTH Aachen, Germany; and Correspondence: *corresponding author, h.fischer{at}rwth-aachen.de
The strength of ceramic restorations depends on the occlusal surface roughness of the veneering porcelain, which is influenced by the final preparation. The hypothesis of the study was that roughnesses below a critical microscopic defect size—based only on fracture mechanics considerations—also affect flexural strength. The bending failure stress was evaluated on standard specimens of 4 veneer ceramics with 4 different surfaces of defined roughnesses, respectively. A linear correlation was found between roughness and failure stress. A "roughness-free" failure stress value was predicted for each tested material. This theoretical value can represent the "true" strength of the respective ceramic material. We conclude from our results that the final preparation of a ceramic restoration is critical to the strength of the material, and that ceramic veneering materials can be compared more objectively with respect to their strength by means of roughness-free strength values.
Key Words: dental ceramics • surface roughness • failure stress • roughness-free strength
One problematic aspect of ceramic materials is the large scatter in strength (Green, 1998; Munz and Fett, 1999). The scatter in strength is caused by the microscopic defect population, which is characteristic of brittle materials. The microscopic defects are statistically distributed due to kind, form, orientation, position, and size. Therefore, ceramics have a low mechanical reliability, since no exact failure limit can be defined. Only a statistical probability of failure can be given (Ritter, 1995). A parameter which affects the flexural strength of a ceramic component, besides the natural microscopic defect population, is surface roughness. The surface roughness of a ceramic restoration depends on the final preparation of the veneering material. This is important, since almost every dental restoration is selectively ground to correct static and dynamic occlusal interferences (Denbo, 1990; Olthoff et al., 2000; Olsson and Lindqvist, 2002). If the peak-to-valley height of the surface roughness is in the range of the critical defect size value, the roughness can affect the flexural strength. The fundamental fracture mechanics equation (Griffith criterion) is given by
where ac is the critical defect size, The critical defect size can be estimated if the (flexural) failure stress of a brittle material is known. Based on the equation, there should theoretically be no risk of failure, if the specimen consists only of microscopic defects below this estimated critical value. Our hypothesis is that surface roughnesses below this microscopic defect size—based only on fracture mechanics considerations—also affect the flexural strength distribution of ceramic materials. We created different defined surface roughnesses on flexural strength specimens to prove this hypothesis. The strength was evaluated in a four-point bending test arrangement. Based on these results, the correlation between surface roughness and failure stress was evaluated. We also discuss the significance of these results for the dental field.
Standard specimens (1.5 x 3.0 x 30.0 mm) were manufactured, according to the company’s recommendations, from the veneering materials of the following dental ceramic systems: Empress 1 (n = 80) and Empress 2 (n = 80) (both IvoclarVivadent, Schaan, Liechtenstein), Symbio Ceram (n = 80) (Degussa Dental, Rosbach, Germany), and Vita Akzent (n = 80) (Vita Zahnfabrik, Bad Säckingen, Germany). The specimens of each material were divided into 4 batches, each consisting of n = 20 specimens. One batch was polished on a rotation ground machine (Typ AW-10; Exakt Apparatebau, Norderstedt, Germany). Silicon nitride ground papers (grits 800, 1200, and 2400; Struers, Willich, Germany) were used for the grinding process. This batch is called ’Polished’ in this study. The other 3 specimen batches of each material were air-abraded (machine Topstar; Bego, Bremen, Germany) with blasting materials of different grain sizes. The specimens of batch 2, called ’Korox 50’, were air-abraded at 4 bars (= 4•105 Pa) with 50 µm, batch 3 (’Korox 110’) at 3 bar with 110 µm, and batch 4 (’Korox 250’) at 2 bar with 250 µm (mean size) blasting material (99.6% aluminum oxide; Bego, Bremen, Germany). The different air-abrading (sandblasting) processes and the polishing process with silicon nitride ground papers were performed only to create different surfaces of defined roughnesses, and not to simulate different clinical treatment options. Subsequently, all specimens were annealed at a temperature 100°K below the respective transition temperature of the material in a ceramic oven (Austromat 300; Dekema, Freilassing, Germany) for 10 hrs to minimize any residual stresses induced by the ceramic process or by the surface treatment (air-abrading and polishing). After the annealing process, the specimens were very slowly cooled to room temperature in the closed oven overnight. The surface roughness of all specimens was examined by means of a stylus analyzer (Perthometer PRK/S6P; Feinprüf Perthen, Göttingen, Germany). The roughness values Ra, Rz, and Rmax were evaluated on those sides of the specimens which were loaded by maximum tensile stresses during the subsequent flexural strength test between the inner roller spans of the four-point bending test arrangement. The value Rmax was taken as a criterion for judging the influence of the roughness on the strength. This value is defined as the maximum of the peak-to-valley heights of the measured (reference) section.
The flexural strength values were evaluated in a four-point bending test arrangement on a universal testing machine (Z030; Zwick, Ulm, Germany) according to DIN EN 843-1 (1995). The outer and inner roller spans were 24 mm and 12 mm, respectively. The crosshead speed was 0.5 mm/min. The mean strength values and their standard deviations were determined. Moreover, the characteristic strength values
The roughness values for Ra were in the range of 0.3 to 6.1 µm, for Rz in the range of 1.1 to 33.7 µm, and for Rmax in the range of 2.5 to 45.2 µm (Table 1  ). The specimens of the ’Polished’ batch showed the smoothest, and those of the ’Korox 250’ batch the roughest surfaces, respectively. The mean failure stress values were in the range of 65.3 to 102.7 MPa (characteristic strength: 68.3 to 108.3 MPa) for Empress 1, 48.1 to 88.0 MPa (50.5 to 93.7 MPa) for Empress 2, 49.2 to 72.9 MPa (52.7 to 77.8 MPa) for Symbio Ceram, and 48.6 to 72.4 MPa (52.0 to 77.2 MPa) for Vita Akzent. A linear correlation between the roughness values Rmax and the mean failure stress values  m, as well as between the roughness values Rmax and the characteristic strength values 0, was found for all 4 investigated materials (Fig.). The correlation coefficients for the linear regression curves (least-square fits) were in the range of 0.9796 to 0.9998 for the analysis based on a Gauss stress distribution, and between 0.9725 and 0.9923 based on a Weibull stress distribution (Table 2).
The analysis of the effect of the surface roughness of a ceramic veneering material on failure stress is important, since almost every dental restoration is selectively ground chair-side to avoid occlusal interferences. Stepwise increased roughnesses were created on the surfaces of the specimens. The ’Korox 250’ condition was the roughest, and the ’Polished’ condition the smoothest. The mean failure stress values based on a Gauss stress distribution, as well as the respective characteristic strength values based on a Weibull stress distribution, can be correlated directly to the roughness parameters Ra, Rz, and Rmax for all 4 tested ceramics. The rougher the surface, the lower was the flexural strength of the specimens, without exception (Fig.  ). These results were not influenced by residual stresses, since the specimens were carefully annealed after the surface treatments. The strength of the specimens decreased continuously with the increase in surface roughness. This finding is in good agreement with those of another study (de Jager et al., 2000), in which this correlation was also found on dental ceramics, although the mean surface roughness value Ra was taken as the criterion for describing the surface condition in that study. We decided to use the Rmax value as a criterion, considering the ’weakest-link principle’ of the failure behavior of ceramic materials. Moreover, in the study mentioned, the strength was evaluated in a biaxial test, which induces a very complex state of stress in the specimens. The four-point bending test arrangement induces a much more homogenous state of stress.
The correlation coefficients of the least-squares fits through the respective data points (Fig.
The value Rmax is the roughness parameter that indicates the magnitude of the peak-to-valley height introduced by the surface treatment. It varied between 2.5 and 45.2 µm (Table 1
The authors thank Mr. Franz Jungwirth for his help with the flexural strength measurements and the dental companies for supplying the ceramic materials. This work was funded by the LuFG Zahnärztliche Werkstoffkunde/RWTH Aachen (internal fund). Received for publication February 3, 2003. Revision received July 7, 2003. Accepted for publication September 8, 2003.
Journal of Dental Research, Vol. 82, No. 12,
972-975 (2003)
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
|
|
 |
|
Sign In to gain access to subscriptions and/or personal tools.
TOP
ABSTRACT
INTRODUCTION
