Die Rolle der Superkompatibilität auf die Ermüdung von Formgedächtnislegierungen
Zusammenfassung der Projektergebnisse
The key results of the project are summarized in Fig. 13. The visualization (Fig.13a) of the cofactor conditions [34] is based solely on freestanding sputtered films produced in our lab. Figure 13 displays the cofactor conditions for the di`erent composition fabricated. In (b) the remanent strain after 100 cycles is depicted. The thermal hysteresis and stress hysteresis in dependence of λ2is depicted in (c) and (d). Note that the remanent strain for the Ti51.2Ni36.5Pd12.3(*) is taken after 91 cycles instead of 100 cycles due to early breakdown. This type of plot allows to visualize directly the deviation of the fulfilment of the cofactor conditions. Here TiNiPd-based shape memory alloys fulfill the cofactor conditions for type I/II twins the least and TiNiCu/Co best, while TiNiCuPd is in between, in this type of projection. Compared to TiNiCu, the compositional sensitivity is greatly enhanced and compatibility decreased notably under applied stress. In addition (not shown) we found that the difference between the CCII values for compound and type I/II twins is larger in TiNiPd, followed by TiNiCuPd, TiNiCu and TiNiCuCo. Typically, compound twins are preferred when λ₂ < 1 and type I/II twins when λ₂ > 1. Since all types of twins are typically observed, we assume that a smaller overall difference between these CCII may enhance the flexibility of nucleation and growth of low energy interfaces. Our results for Ti-poor compositions suggest that fulfilling the CCII improves cyclic stability during stress-induced transformations. However, we were unable to fabricate TiNiPd films with grain sizes comparable to those of TiNiCuCo films, which are necessary for achieving ultra-low fatigue conditions. Satisfying the cofactor conditions alone is insufficient to produce low fatigue polycrystalline materials. Minor deviations from ideal compatibility can be compensated for by grain refinement, particularly in thermally induced transformations. A positive effect of λ₂ in TiNiPdFe based alloys was observed when other microstructural influences (precipitation hardening) were minimized. The thermal hysteresis depends mainly on λ₂ and is strongly influenced by the precipitation phase content. We find that Ti-rich compositions, when heat treated at high temperatures typically exhibit 3 to 4 °C higher thermal hysteresis than their Ti-poor counterparts, as twinning processes are hindered. Stress hysteresis follows the same trend but can be reduced substantially through low temperature annealing. The effect on thermal hysteresis is likely given, but hard to estimate as the latent heat is reduced not allowing for a precise transformation temperature determination. Low temperature annealing promotes a reversible continuous transformation from austenite to martensite as the energy barrier reduces and a high number of nucleation sites for martensite is formed by stress fields of coherent intergranular Ti2Pd,Cu plate precipitates. They substantially increase the resistance to functional fatigue. As the lattice mismatch of the Ti2Pd,Cu precipitates are further reduced, the thermal induced martensitic transition is subsequently suppressed, resulting in a partial strain-glass transition and embrittlement for Ti53.6Ni36.2Pd11.2. While the differences above were known partly prior to the start of the project, we underestimated the compositional sensitivity of TiNiPd SMAs on the phase transformation behavior. Overall, we found 5 factors leading to the significant difference in material behavior compared to TiNiCu. • TiNiPd alloys exhibit a strong variation in crystallographic compatibility within ±3 at.% Pd, whereas TiNiCu alloys maintains a relatively stable compatibility across broader Cu concentration ranges. • The transformation temperatures of TiNiPd are highly sensitive to Ti and Pd contents, particularly below 50 at.% Ti. In contrast, TiNiCu alloys show moderate influence of transformation behavior by the Ti and Cu content. • The mismatch of Ti2Pd(Ni) is smaller compared to Ti2Cu(Ni), affecting the nucleation and growth of martensite and potentially promoting strain glass formation together with the compositional dependency. • TiNiPd films consistently exhibit larger grain sizes than TiNiCu alloys, relatively independent of Ti content. Attempts to reduce grain size via Co or Ni multilayers were not successful. • Residual austenite or R-phase was present in all TiNiPd-based alloys, regardless of heat treatment and composition. Residual austenite was enhanced in low temperature heat treatments, a phenomenon being absent in TiNiCu-based systems. While significant progress has been made in understanding the key factors for achieving high functional stability and shape memory behavior of TiNiPd and TiNiCu based shape memory alloys, several new research questions arise during the project. The mechanism and phases responsible in the formation of retained austenite / R-phase and a potential strain glass are not fully understood. Here temperature dependent TEM imaging of Ti-lean and Ti-rich TiNiPd alloys heat treated at low and high temperatures should be conducted. The potential strain glass behavior should be further analyzed through comparative studies of low temperature Ti-rich TiNiPd alloys and Ti50NiPd films which did not exhibit a phase transition during our studies. We found that the small misfit of Ti2Pd(Ni) precipitates leads to stronger suppression of martensitic transition, increased transformation stress and brittleness. Their impact on martensitic transformation at higher Pd concentrations, particularly considering increased lattice misfit, λ₂ deviations and reduced grain sizes we found with increased Pd content (Pd~20 at%). An extended microstructure analysis using temperature and stress dependent synchrotron measurements beyond the current experimental limits particularly at temperatures below -100 °C and for materials with λ₂ > 1 and λ₂ < 1, with and without plate precipitates and varying precipitate sizes would provide further insights into the exact microstructure of the involved phases and the underlying mechanisms, especially martensite variant selection, the crystallographic compatibility and potential additional phase transition causing the sloped stress transformation plateau. To achieve this goal, a broader interdisciplinary collaboration is essential combining at least expertise in theoretical modeling, advanced microstructural characterization and the fabrication of shape memory alloys. During the project, it became evident that TiNiPd based alloys did not have the required functional stability for actuator or elastocaloric applications compared to TiNiCu and TiNiCuCo alloys. We have therefore focused on TiNiCu [D] and TiNiCuCo [E] alloys in application-oriented collaborations, which has resulted in further publications from the project.
Projektbezogene Publikationen (Auswahl)
-
Cu-rich Ti52.8Ni22.2Cu22.5Co2.5 shape memory alloy films with ultra-low fatigue for elastocaloric applications. Journal of Applied Physics, 127(22).
Bumke, Lars; Zamponi, Christiane; Jetter, Justin & Quandt, Eckhard
-
Coherent Precipitates as a Condition for Ultra-Low Fatigue in Cu-Rich Ti53.7Ni24.7Cu21.6 Shape Memory Alloys. Shape Memory and Superelasticity, 7(4), 526-540.
Bumke, L.; Wolff, N.; Chluba, C.; Dankwort, T.; Kienle, L. & Quandt, E.
-
Bistable Actuation Based on Antagonistic Buckling SMA Beams. Actuators, 12(11), 422.
Chen, Xi; Bumke, Lars; Quandt, Eckhard & Kohl, Manfred
-
Effect of Ti2Pd(Ni) on the Transformation Behavior in Sputtered Ti-rich TiNiPd Shape Memory Alloys
L. Bumke, N. Wolff, L. Kienle & E. Quandt
