An investigation of the material and model parameters for a constitutive model for MSMAs

A two dimensional constitutive model capable of predicting the magneto-mechanical response of a magnetic shape memory alloy (MSMA) has been developed and calibrated using a zero field-variable stress test¹. This calibration approach is easy to perform and facilitates a faster evaluation of the three calibration constants required by the model (vs. five calibration constants required by previous models^2,3). The calibration constants generated with this approach facilitate good model predictions of constant field-variable stress tests, for a wide range of loading conditions¹. However, the same calibration constants yield less accurate model predictions for constant stress-variable field tests. Deployment of a separate calibration method for this type of loading, using a varying field-zero stress calibration test, also didn't lead to improved model predictions of this loading case. As a result, a sensitivity analysis was performed on most model and material parameters to identify which of them may influence model predictions the most, in both types of loading conditions. The sensitivity analysis revealed that changing most of these parameters did not improve model predictions for all loading types. Only the anisotropy coefficient was found to improve significantly field controlled model predictions and slightly worsen model predictions for stress controlled cases. This suggests that either the value of the anisotropy coefficient (which is provided by the manufacturer) is not accurate, or that the model is missing features associated with the magnetic energy of the material.