Publication Name

Acta Materialia

Abstract

Stress-induced martensite formation is the predominant mechanism during the deformation of various functional and structural Ti- alloys. To predict the performance of these alloys, stress-induced martensite formation was modeled as a function of crystal orientation and stress state. We present an integrated micromechanical modeling approach using finite element (FE) analysis and an elastic spectral solver based on Fast Fourier Transforms (EFFT), which allows direct correlation of the available work from martensite formation under complex stress-states in an in-situ characterized microstructure. The model is applied as a virtual analogue of an experimental 3-point bending test of a metastable β Ti–10V–2Fe–3Al alloy containing 5% α. The FE model incorporates the experimental β microstructure from electron backscattering diffraction (EBSD) with anisotropic elastic behavior. The EFFT solver uses strain fields in two local regions from the FE model to predict local stresses in experimental microstructures containing β and α phases. The stress and orientation data are used to predict the available work from stress-induced α″ martensite formation of the six different α″ variants. It was found that stress and available work concentrated around the tips of lamellar α, making these regions preferred nucleation sites for α″ formation, which is in excellent agreement with the in-situ experimental observations. This suggests that the local stress conditions in the β phase play a more important role in triggering α″ formation than the compositional inhomogeneity in the β phase at α-β interfaces. Using the integrated model, the first variants forming at low levels of deformation could be confidently predicted. However, the transformation behavior at elevated levels of deformation could not be properly captured. This indicates the need of further model improvement, including Dirichlet boundary conditions for the EFFT solver and accounting for the role of plasticity induced by phase transformation.

Volume

240

Article Number

118342

Funding Number

DP170100836

Funding Sponsor

Australian Research Council

Share

COinS
 

Link to publisher version (DOI)

http://dx.doi.org/10.1016/j.actamat.2022.118342