Modelling of the cyclic behaviour of superalloys

Licentiate thesis, 2013

In the aviation industry today there is a trend of increasing operating temperatures in turbine engines. The advantages with higher temperatures in the engines are improved performance and efficiency. During a flight cycle the engine components are exposed to cyclically varying thermal and mechanical loads, which might cause thermo-mechanical fatigue (TMF) of the material. For prediction of the TMF life it is important to have a material model that can cover the temperature range of interest for the analysed component. The material model should also capture the decisive material mechanisms active in the component during the flight. Therefore the constitutive model should be able to mimic phenomena such as cyclic hardening/softening, the Bauschinger effect, ratchetting, shake down, creep and stress relaxation. The objective of this thesis is to investigate and develop modelling of cyclic plasticity phenomena observed for superalloys in low-cycle fatigue (LCF) experiments at different temperatures. In the first paper, the mechanical behaviour of the Ni-based superalloy Haynes 282 is studied at room temperature and 650℃. A Chaboche-type of plasticity model is chosen as a base model and calibrated against LCF experiments. The robustness and uniqueness of the identified material parameters are ensured by performing sensitivity analyses. The plasticity model is extended to include several kinematic hardening variables and it is studied how this affects its response. Furthermore, the influence of uncertainties in experimental data on identified material parameters, fatigue life predictions and finite element (FE) predictions is investigated. The modelling of the cyclic behaviour of Haynes 282 is continued in the second paper where the base model is modified to account for the cyclic softening observed at high temperatures. The material model is calibrated for a range of elevated temperatures and a temperature dependence is established for the material parameters. The temperature dependence is validated against experiments with good results. In addition, an FE example is given to illustrate the consequences when including cyclic softening in the material model. The slow evolution of the cyclic softening requires many loading cycles to develop and therefore a technique for cycle extrapolation is incorporated in the FE analysis to increase the efficiency of the computations. In the third paper, the base model is used to capture the initial mean stress relaxation observed for a high strength Titanium alloy. The calibrated material model is included in FE analyses of LCF crack growth experiments. Correlation between experimental and FE results indicates a potential for the prediction of crack length based on the measured load drop during LCF testing.