ICF10B

EXPERIMENTAL PROCEDURE The specimen is shown in Fig.1. The material used for the experiment was PZT-5, which has a tetragonal crystal structure at room temperature and an average 3 mµ grain size. Specimens were cut and polished to dimensions of . Gold electrodes were sputtered onto the upper and lower surfaces of the PZT-5 specimen. The specimens were poled in the direction perpendicular to the crack faces under an electric field of 3 (the coercive field, ( ) 40 6 3 mm × × cE E V mm c / 1000 = ) at a temperature of 130 . At first, a short notch with a width of 0.25mm was generated on the specimens. Then, by use of a Combined Load Device that was specially designed for producing pre-crack, the specimens were pre-cracked. For the details of this special technique, one can read the paper of Zhang et al [9]. The specimen was immerged in a silicon oil container that was made of transparent and insulating plexiglass and subjected to a cyclic electric field through the bolts that attach to the electrodes. The sine wave forms were applied to the specimens under a wide amplitude range of 0.8E Co c~3.2Ec. Because the specimen was plated on the upper and lower surfaces, if no crack existed in the specimen, the electric field in the specimen would be uniform. The power used in experiments was a voltage-adjustable alternating high voltage supplier, which can reach its highest output of 30 kV. There are five available frequencies which are 50 Hz, 4Hz, 3Hz, 2Hz and 1Hz. ± 6 40 sP Electrode Initial crack Unit (mm) Dent 3 Fig. 1: Schematic of the specimen RESULTS AND DISCUSSIONS All specimens were observed through microscopy with 400 times amplification before loading. No obvious microcrack was found and the crack tips were very sharp. Fig. 2 exhibits that a crack propagates from an initial penetrated crack in the specimen. The experimental results [3] and the theoretical explanation [10] showed that there exists the anisotropy of fracture toughness in ferroelectrics. That is, the crack propagates along the direction perpendicular to the electric field more easily than along the direction parallel to the electric field. The reason that the main crack grows along the middle crack surface is due to the fact that the crack in the middle surface would bring the greatest disturbance for the electric and mechanical field. This may cause the strongest electric field and stress concentrations that induce the largest energy release rate. Fig. 2: The propagated macroscopic crack in the specimen In the test, it was also found that the crack did not grow at once when the electric field was applied. It always frizzed for some time, then propagated quickly, especially under a low electric field. Moreover, there were some phenomena that occur before the crack grows. For example, some little bubbles produced from the main crack and the buzz noise can be heard. When the loading frequency was low, it can be found that sparks occurred inside the crack, and some black powder appeared on the crack faces. After a period of fatigue cycling, detached the specimen along the crack, we found that the crack surface generated by electric-field-induced fatigue was black. Note that the buzz always occurred and was independent of the crack length that would influence resonant frequency. Since the electric sparking was observed, it can be assumed that the electric breakdown occurred inside the crack. According the theory on electric breakdown [11], if there are bubbles in the liquid, since the electric fields in dielectrics under alternating current (AC) is inversely proportional to their permittivities, the electric field in the bubbles is always higher than that in the liquid. Because the electric field of electric breakdown in bubbles is less than that for liquid, the ionization

RkJQdWJsaXNoZXIy MjM0NDE=