赝势
密度泛函理论
材料科学
堆积
叠加断层
Atom(片上系统)
层错能
分子物理学
凝聚态物理
结晶学
计算化学
位错
物理
化学
核磁共振
计算机科学
嵌入式系统
复合材料
作者
Magali Benoit,Nathalie Tarrat,Joseph Morillo
标识
DOI:10.1088/0965-0393/21/1/015009
摘要
Bulk properties of hcp-Ti, relevant for the description of dislocations, such as elastic constants, stacking faults and γ-surface, are computed using density functional theory (DFT) and two central force embedded atom interaction models (Zope and Mishin 2003 Phys. Rev. B 68 024102, Hammerschmidt et al 2005 Phys. Rev. B 71 205409). The results are compared with previously published calculations, except pair potential calculations, which are not appropriate for the description of the metallic bond. The comparison includes N-body central force (NB-CF) and N-body angular (NB-A) empirical potentials, tight-binding approximation to the electronic structure (TB), DFT pseudopotential (DFT-P) and all electron (DFT-A) calculations. None of the considered interaction models are fully satisfactory for the description of these properties. In particular, NB-CF, NB-A and TB interaction models are unable to describe the softening of the easy prismatic γ-surface leading to the appearance of a metastable stacking fault, as evidenced in all the DFT calculations. Most often, when the basal stacking fault excess energy is underevaluated, this leads to an inversion of the energetic stability between the I2 basal and the prismatic easy stacking faults. Even the DFT-pseudopotential calculations need to be improved regarding the description of the shear elastic constants. The implications of these results on the core structure and gliding properties of the screw dislocation are analyzed. The calculated dissociation lengths into Shockley partials in both the basal and prismatic planes for the different models compare well with the measured ones in the corresponding simulations of the dislocation core structure when available. Finally, the Peierls stress is also evaluated using the Peierls–Nabarro model and compared with the experimentally measured one.
科研通智能强力驱动
Strongly Powered by AbleSci AI