DYNAMIC FATIGUE OF FELDSPATHIC PORCELAIN

Several studies (Sherrill and O'Brien, 1974; Southern and Jorgensen, 1 974; Jones, 1983) have shown that stress corrosion fatigue occurs in d ental porcelains. Morena et al. (1986) reported on an assessment of sl ow crack growth parameters for dental ceramics. The purpose of the stu dy reported here was to evaluate the fatigue parameters of a model exp erimental porcelain using dynamic fatigue testing. This test procedure makes use of several constant stressing rates to perform strength tes ts. Dynamic stress testing was first described by Evans (1974) and lat er defined as a distinct test modality by Ritter (1978). From such dat a, the fatigue parameters can be calculated. Theses fatigue parameters , n and sigma(fo), are, respectively, the crack growth exponent from t he crack velocity expression and a materials constant which is depende nt on the test environment and the inert (moisture-free) strength. The model porcelain was made from 60% component 1 and 40% component 3 acc ording to the Weinstein patent (Weinstein, et al., 1962). The biaxial flexure strength of 300 specimens 1 mm thick was tested in 37 degrees C water by testing 50 samples at each of 6 constant stressing rates: 1 00, 10, 1, 0.1, 0.01, and 0.001 MPa/s. One hundred specimens were test ed in a moisture-free environment at 100 MPa/s using a servo-mechanica l testing machine. A commercial porcelain (Jelenko Gingival-Lot# 2012, Jelenko Dental Health Products, Armonk, NY, USA) was chosen as a refe rence material. One hundred twenty specimens were tested using the sam e procedures as those used for the model porcelain; however, only 20 s amples were tested for 5 stressing rate groups and an inert group. Fat igue parameters with their respective standard errors for the model po rcelain were n=28 +/- 1.00, and sigma(fo) = 58.7 +/- 0.31 MPa. Values for the commercial porcelain were n=27.6 +/- 2.00 and sigma(fo) = 49.1 +/- 0.51 MPa. These parametric values were obtained from the constant s of a linear regression of LOG failure stress vs LOG time to failure obtained from constant stressing rate data for all of the stressing ra te groups tested in water for each porcelain. Readers are cautioned th at direct extrapolation of the results of this research to clinical ap plications could be misleading. The additional effects of porcelain to metal bonding on fatigue are poorly understood at this time. A report on preliminary results of porcelain-fused-to-metal disk strength has been presented (Fairhurst et al., 1992), and fatigue parameter evaluat ions at this laboratory are in progress.