| Summary: | Bulk shape memory ceramics (SMCs) are attractive for their high transformation temperatures and transformation stresses compared to shape memory metals but exhibit transformation-induced cracking due to mismatch stresses arising at the grain boundaries. Current strategies for mitigating cracking in SMCs involve moving towards smaller volume structures, which feature high surface area for stress relaxation and fewer grain boundaries to minimize transformation stresses. While this approach has proven successful, it typically limits SMCs to sample sizes at the micrometer-scale in micro-pillar and micro-particle structures. Here it is proposed that the strategy of simply lowering the grain boundary area can result in bulk SMCs that do not crack as these structures, like the micropillars and microparticles, satisfy the constraint of having relatively few sites of high stress concentrations. This led to the investigation of optical floating zone and cold crucible induction crystal growth of zirconia-based SMCs. Single crystals produced through the cold crucible induction melting route were found to be of high quality and ultimately increased the number of thermally induced martensitic transformation cycles from 5 in bulk materials out to at least 125.
The tetragonal-to-monoclinic transformation behavior of yttria-doped zirconia in polycrystalline and single crystalline forms were compared over many transformation cycles collected in the differential scanning calorimeter. The evolution of thermal hysteresis and transformation strains were used to characterize thermal cycling performance of each structure. Whereas single crystals had very repeatable transformation behavior in terms of hysteresis and strain amplitude, polycrystals degraded dramatically as they accumulate cracking damage with repeated cycling. As the polycrystal evolved from a pellet to granular packing of loose single crystals/grains, the energy dissipation converged with that of the single-crystal structure, and the energy spent on cracking throughout that process is captured by calorimetry analysis.
We then explored grain size effects in cyclic martensitic transformations in polycrystalline structures to study cracking-induced disaggregation as a function of grain size from ~0.6 to 7.9 µm. A smaller grain size was found to increase the number of cycles required to disaggregate the pellet because of the larger amount of grain boundary area that must crack. Calorimetry analysis showed that the energy relieved through cracking decreases with increasing grain size and suggested an apparent material length scale of ~2 micrometers for the stress relief zone. By comparing data from initial and final transformation cycles, grain size and particle size effects could be developed and compared to relationships already established in shape memory alloys.
These results all verify that grain boundaries play a key role in damage accumulation/ evolution during cyclic martensitic transformation and that microstructural control can extend the size-scale of viable single crystal or oligocrystal SMCs from the micro- to the millimeter scale
|
|---|