|Project manager:||Prof. Dr. rer. nat. Gerold A. Schneider|
|Projekt worker:||Dr. rer. nat. Hans Jelitto|
|Supported by:||DFG, Grenzflächenbruch in Piezoelektrika (Projekt SCHN 372/11-2)|
Technische Universität Dresden, Institut für Festkörpermechanik, Germany
In the past different methods have been developed for fracture experiments to measure R-curves with stable crack growth. In the theoretical research the search for a general valid fracture criterion is still a challenging issue. For pure mechanical loading the situation has been understood well. If the stress intensity factor KI or the energy release rate G reaches the critical values KIC or GC, respectively, the crack propagates. With an additional electric load and especially when using piezoelectric materials, the situation becomes more complicated. In this case a general valid fracture criterion for all loading conditions has not been found up to now.
Thus, a very stiff testing machine has been developed and optimized in order to enable stable crack growth in 4-point-bending including electrical loads (see Fig. 1). The device allows simultaneous measurement of the R-curve (fracture toughness) and measurement of all energies contributing to the total energy release rate. These contributions are determined with the special small signal compliance method, where the changes of the mechanical compliance, the piezoelectric compliance, and the capacitance during crack advance are measured simultaneously.
Figure 1: Stiff testing machine for stable crack growth with electronics for high voltage, for the modulation of mechanical and electric loads, as well as for data acquisition (left). Rigid frame with the mechanical arrange-ment inside (right).
Interfacial stable fracture between metal and ferroelectric ceramics is investigated under electromechanical loading. The main focus is put on the material response under mechanical mixed mode loading, e.g. a combination of mode I and II (Fig. 2). The 4-point-bending specimens are prepared from commercial multilayer actuators. The aim is to measure fracture curves and critical energy release rates as well as to determine the crack morphology dependent on different materials and load conditions. These experi-ments are valuable for the theoretical research and are of vital interest for the industrial application.
Figure 2: Sample and support rollers for different asymmetrical load arrangements. a) and b) mixed mode loading (KI and KII) c) pure KII-loading. The inner metal electrodes in the sample have distances of 90 µm and are oriented parallel to the crack surface.
Ferroelectric PZT bulk material, which is poled mechanically, is slowly fractured with different poling directions. The mechanical poling is achieved by pressing with 300 MPa. Compared to the electric poling the mechanical poling has the advantage that the results can be predicted more easily by the theory, because there is no effective piezoelectric coupling. Controlled fracture experiments with pure mechanical loading and different poling geometries have been performed already. Hereby, a significant influence of the poling direction on the fracture toughness has been observed. Measurements with additional electric loading are planned next.
Stable crack advance is measured in structural micro-laminate ceramics, which are produced at the EMPA in Zürich, Switzerland. The ceramic, consisting of Si3N4- and Si3N4/TiN-layers, have alternately internal tension and compression stress, yielding an increasing apparent fracture toughness. Up to now two different layer geometries have been tested. The maximum fracture toughness of about 18 MPa*m1/2 is quite high. Therefore an extreme stiff bending device is needed. We use the custom made testing machine, developed at the TUHH (Fig. 1), as well as a commercial version (EXAKT GmbH, Norderstedt). Figure 3 shows an R-curve measured with the EXAKT-machine with stepwise increasing fracture toughness.
Figure 3: Measured R-curve in a ceramic with specifically designed micro layers (without electric load). The continuous line represents the model calculation. The measurement was performed with a single specimen.