A computational mesoscale evaluation of material characteristics of porous shape memory alloys

Computational techniques are used to estimate porous shape memory alloy material behavior in this study. Mesoscale analysis techniques based on a two-dimensional finite element method are used to evaluate a representative volume element (RVE) to determine effective material response. The porous shape memory alloy is defined as comprised of two constituents, dense NiTi shape memory alloy (SMA), and pores that are defined as free of all material and distributed randomly. An existing rate-independent type constitutive model is used to define the dense SMA material behavior. Open porosity is considered with the axis of generation perpendicular to the plane of the two-dimensional RVE. Pores are allowed to vary in size and shape with no limit on maximum pore size. Computational evaluations are completed for porous SMA for a range of pore volume fractions. Effective material behavior is calculated and results are compared with computationally determined material behavior based on the micromechanical unit cell finite element method. Calculated effective material elastic and transformation characteristics compare well between the two methods for low to moderate (less than or equal to0.3) pore volume fractions. Significant variations are seen for higher pore volume fractions (>0.3) due to spatial orientation and percolation effects that are not considered in unit cell analyses.