A COMPUTATIONAL MICROMECHANICS STUDY ON VARIANT-COALESCENCE IN A CU-AL-NI SHAPE-MEMORY ALLOY

In shape memory alloys (SMA) reorientation of crystallographic variant s upon loading takes place in a temperature regime lower than the mart ensite finish temperature M(f) a process which is denoted as variant c oalescence (VC). The macroscopic deformation behavior due to VC of a p olycrystalline Cu-Al-Ni shape memory alloy is studied by introducing a combined computational micromechanics and continuum-thermodynamics fr amework. For the simulation it is assumed that the VC starts from a se lf-accommodating morphology of thermo-induced martensite. It proceeds until each of the diamond-shaped plate groups contains the most favora ble of its four variants. The kinematics of the parent --> martensite transformation are described by stress-free transformation tensors for crystallographic variants of thermo-induced martensite emerging along habit planes. We persue a thermodynamic field concept for VC by intro ducing a Gibbs free energy formulation to describe the thermodynamic s tate of the stressed polycrystalline mesodomain. We prescribe the kine tics of the reorientation process on the size scale of the crystallite s in terms of the mobility of the boundaries between martensitic Varia nts (intervariant boundaries). Then the assumed reorientation process is the subject of a micromechanical simulation. The transformation-ind uced microstress and microstrain fields are computed. By checking the thermodynamic admissibility of any small increment of the reorientatio n process the corresponding required increment of externally-applied t ensile stress can be calculated and the overall stress-strain curve fo r the assumed reorientation process results. VC is studied in the two limit cases of high and low mobility of the intervariant boundaries be tween martensitic variants forming self-accommodating plate groups. Th e effect of the mobility of intervariant boundaries on the stress-stra in curve is quantified by a polycrystalline finite-element analysis.