Three Dimensional Grain Boundary Modeling in Polycrystalline Plasticity


Creative Commons License

YALÇINKAYA T. , ÖZDEMİR İ., Firat A. O.

21st International ESAFORM Conference on Material Forming (ESAFORM), Palermo, İtalya, 23 - 25 Nisan 2018, cilt.1960 identifier identifier

  • Cilt numarası: 1960
  • Doi Numarası: 10.1063/1.5035076
  • Basıldığı Şehir: Palermo
  • Basıldığı Ülke: İtalya

Özet

At grain scale, polycrystalline materials develop heterogeneous plastic deformation fields, localizations and stress concentrations due to variation of grain orientations, geometries and defects. Development of inter-granular stresses due to misorientation are crucial for a range of grain boundary (GB) related failure mechanisms, such as stress corrosion cracking (SCC) and fatigue cracking. Local crystal plasticity finite element modelling of polycrystalline metals at micron scale results in stress jumps at the grain boundaries. Moreover, the concepts such as the transmission of dislocations between grains and strength of the grain boundaries are not included in the modelling. The higher order strain gradient crystal plasticity modelling approaches offer the possibility of defining grain boundary conditions. However, these conditions are mostly not dependent on misorientation of grains and can define only extreme cases. For a proper definition of grain boundary behavior in plasticity, a model for grain boundary behavior should be incorporated into the plasticity framework. In this context, a particular grain boundary model ([1]) is incorporated into a strain gradient crystal plasticity framework ([2]). In a 3-D setting, both bulk and grain boundary models are implemented as user defined elements in Abaqus. The strain gradient crystal plasticity model works in the bulk elements and considers displacements and plastic slips as degree of freedoms. Interface elements model the plastic slip behavior, yet they do not possess any kind of mechanical cohesive behavior. The physical aspects of grain boundaries and the performance of the model are addressed through numerical examples.