Search for compatible Zirconia-based shape memory ceramics

点击翻译按钮获取中文摘要

Metallic shape memory alloys (SMA) show two properties of interest: the shape memory effect used for solid state actuators and the superelasticity e.g. for self-expanding medical implants. They are based on a reversible, thermal or stress-induced phase transition of the first order between austenite and martensite phases with large internal stresses, which lead to large strains and enthalpy changes. The reversibility is achieved through the compatibility of the two phases, which leads to a reduction in the local stresses at the interface and the hysteresis. This compatibility can be expressed by the condition that the mean eigenvector of the transformation matrix ideally has to be one.Due to the elements used in metallic SMA, use under corrosive chemical conditions or at extreme temperatures is difficult. This motivates a new class of shape memory materials, which consist of metal oxides and are called shape memory ceramics (SMC). Although Swain first identified a suitable oxide (MgO-stabilized ZrO2) in 1986, the lack of compatibility to date leads to the development of cracks and severe fatigue upon cycling.In prior own research the findings regarding compatibility in metallic SMA were transferred to their ceramic counterparts in the system (Y0.5Ta0.5O2)1-x (Zr0.5Hf0.5O2)x. The main results were a significant dependence of the transformation hysteresis on the composition and a strong dependency of the stress-induced orientation of the martensitic variants on the hysteresis, without however approaching the properties of metallic SMA. The dependence on the mean eigenvalue also differs: Given the two probable transformation mechanisms, the results suggest that the composition, in which the compatibility for both transformations is equally well fulfilled, defines the point of lowest hysteresis (equidistance condition).The main goal of this application is the search for SMC with low hysteresis and good shape memory properties. This search is based on a data set containing approximately 70 different ZrO2-based SMC with the transformation temperatures and the crystallographic data. The corresponding working hypotheses are that for SMC the mean eigenvalues ​​of both lattice correspondences must be as close as possible to one and that these conditions lead to two lattice parameter relations that suggest searching for dopants that reduce the tetragonality.In addition to this main objective we will investigate whether a machine learning approach based on the available data set of physical parameters will be appropriate to identify compositions for SMC with significantly improved properties in terms of hysteresis and transformation reversibility.