SUMMARY

The unique functional behaviors such as superelasticity, actuation, and shape memory effect have led to substantial industrial interest in shape memory alloys (SMAs) and motivate the development of models that can describe the complex coupled physics of SMAs from a macroscale and a microscale approach. This work is aimed at addressing two main avenues for modeling development in the broader ICME framework (1) Advancement in robust and accurate physics-based model development that can incorporate the hierarchical mechanisms in material systems while being computationally feasible to predict material response. (2) Leveraging the use of these models to enable material characterisation through calibration of microscale and macroscale models.
Finite strain micromechanical models are better shown to capture the mesoscale coupled physics over phenomenological or small strain formulations by using crystallographic descriptions of the deformation processes. However, all prior models have ignored the rotational component of the deformation gradient for phase transformation (PT) of the variants in their flow rule descriptions, which can lead to significant differences in variant selection and recoverable strain.
This work is particularly aimed at developing a fully-implicit finite-deformation rate-dependent crystal mechanics modeling and computational framework that simulates the coupled elastic-plastic-PT mechanisms in SMA while incorporating the entire deformation gradient for the transformation variants (including the rotational component). This work will further probe and understand the effect of loading path dependency of the final material state for certain load histories and show the differences that arise in including the transformation rotations as well as from the coupled plasticity and PT. Second, the modeling framework will be leveraged to compare and calibrate against X-ray diffraction measurements of martensite variant volume fractions during uniaxial loading of Nickel-Titanium to estimate its material parameters including martensite elastic stiffness constants and the variant interactions using Bayesian inference. Finally, the proposed model will be further utilized to simulate the ground truth SMA response and used to calibrate the macroscale phenomenological model parameters thereby building a multiscale linkage between the microscale and the macroscale response and cutting down cost of carrying out expensive experiments.