TY - JOUR
T1 - Doping effects of point defects in shape memory alloys
AU - Yang, Yuanchao
AU - Xue, Dezhen
AU - Yuan, Ruihao
AU - Zhou, Yumei
AU - Lookman, Turab
AU - Ding, Xiangdong
AU - Ren, Xiaobing
AU - Sun, Jun
N1 - Publisher Copyright:
© 2019
PY - 2019/9/1
Y1 - 2019/9/1
N2 - Doping point defects into shape memory alloys (SMAs) influences their transformation behaviors and mechanical properties. We propose a general Landau free energy model to study doping effects, which only assumes that point defects produce local dilatational stresses coupled to the non-order parameter volumetric strain. Different dopants can be represented by their range of interaction and potency of dilatational stress. Time-dependent simulations based on our model successfully reproduce experimentally observed doping effects in SMAs, including the elevation or suppression of the transformation temperature, the modification of mechanical properties, the appearance of a cross-hatched tweed structure and the emergence of a frozen glassy state with local strain order. We predict that the temperature range for superelasticity will be enhanced in the crossover region between martensite and strain glass. In addition, an Elinvar effect appears most likely in alloys with dopants tending to increase the transformation temperature, which needs to be verified experimentally. Moreover, the two parameters in the Landau model, the interaction range and potency of the dilatational stress, inspire us to identify three material descriptors with which we can construct an empirical machine learning model. The model predicts the transformation temperature, and the slope of the change in transformation temperature as a function of doping composition, enabling an effective search for doped SMAs with targeted properties via machine learning.
AB - Doping point defects into shape memory alloys (SMAs) influences their transformation behaviors and mechanical properties. We propose a general Landau free energy model to study doping effects, which only assumes that point defects produce local dilatational stresses coupled to the non-order parameter volumetric strain. Different dopants can be represented by their range of interaction and potency of dilatational stress. Time-dependent simulations based on our model successfully reproduce experimentally observed doping effects in SMAs, including the elevation or suppression of the transformation temperature, the modification of mechanical properties, the appearance of a cross-hatched tweed structure and the emergence of a frozen glassy state with local strain order. We predict that the temperature range for superelasticity will be enhanced in the crossover region between martensite and strain glass. In addition, an Elinvar effect appears most likely in alloys with dopants tending to increase the transformation temperature, which needs to be verified experimentally. Moreover, the two parameters in the Landau model, the interaction range and potency of the dilatational stress, inspire us to identify three material descriptors with which we can construct an empirical machine learning model. The model predicts the transformation temperature, and the slope of the change in transformation temperature as a function of doping composition, enabling an effective search for doped SMAs with targeted properties via machine learning.
KW - Martensitic transformation
KW - Materials informatics
KW - Strain glass
KW - Time-dependent Ginzburg-Landau simulation
KW - Tweed
UR - http://www.scopus.com/inward/record.url?scp=85068612494&partnerID=8YFLogxK
U2 - 10.1016/j.actamat.2019.06.031
DO - 10.1016/j.actamat.2019.06.031
M3 - 文章
AN - SCOPUS:85068612494
SN - 1359-6454
VL - 176
SP - 177
EP - 188
JO - Acta Materialia
JF - Acta Materialia
ER -