Predicting Fatigue Crack Growth Life in Additive Manufactured Titanium Alloy

  • Xueyuan Wang

    Student thesis: Doctoral ThesisDoctor of Philosophy

    Abstract

    The aim of this PhD project is to investigate the fatigue crack propagation behaviour in an additive manufactured high strength titanium alloy, Ti-6Al-4V. Test specimens were made by the Wire+Arc Additive Manufacture (WAAM) process. The research focus is the fatigue crack behaviour (crack growth rate and trajectory) near interface between WAAM and substrate materials. The challenges are the understanding and assessment of the effects of residual stress, microstructure change and anisotropic material properties as the result of rapid heating and cooling, and rapid solidification cycles in additive manufacturing. This PhD project has focused on numerical modelling and simulation of fatigue crack behaviour. Specimen fabrication and experimental tests were conducted by our collaborators in linked projects. Finite element method (FEM) was employed to evaluate the influence of anisotropic Young's modulus and yield strength properties on the crack tip stress intensity factor and crack tip plasticity. Residual stress distribution in the compact tension, C(T), specimens were obtained by FEM based on experimentally measured residual stress in a much larger WAAM-substrate wall, from which the C(T) specimens were extracted. Residual stress profile was also established by an analytical approach for another WAAM-substrate wall, from which the fatigue crack growth rates (FCGR) for pure WAAM material were measured. Based on these calculations, fatigue crack growth rate and life were predicted by empirical methods from the Linear Elastic Fracture Mechanics, namely the modified Paris law and the Harter T-method. Residual stress effect is accounted for by the superposition method, via the effective R ratio parameter.

    Key findings and main conclusions are: (1) the difference in the stress intensity factor is less than 1% when considering the anisotropic material properties. Therefore, the influence of anisotropic material properties on the crack growth driving force can be neglected. (2) After extracting a C(T) specimen from a larger wall sample, retained residual stress in the C(T) specimen is much reduced; consequently the stress intensity factor due to the residual stress is also small. (3) Residual stress free assumption is not valid for C(T) specimens which provide FCGR data for pure WAAM material. (4) The Harter T-method is better when predicting FCGRs in WAAM material and residual stress effect should not be ignored. (5) Predicted fatigue crack growth life for the specimens containing WAAM-substrate interface have a difference about 25% compared to experiments. Predicted fatigue crack deviation angles for various crack locations and orientations are consistently larger than the experimental measurement, as the crack closure has reduced the effect of mode II stress intensity factor.
    Date of Award2017
    Original languageEnglish
    Awarding Institution
    • Coventry University
    SupervisorXiang Zhang (Supervisor), Michael Fitzpatrick (Supervisor) & Abdul Syed (Supervisor)

    Cite this

    '