The traditional approach to the design of industrial Gas Turbines considers base load operation. This assumption is no longer applicable as owner operators require more operational flexibility and increased availability and reliability. Flexibility in operation manifests as increased cyclic loading and variations in on and off load dwell periods and thermal loads. These complex loading profiles inevitably lead to damage from both creep and fatigue, and interaction of these two damage mechanisms over the duration of the service interval. These interactions can result in higher average creep rates and more damage than expected.
Robust, path dependent modeling approaches are required to better understand the effects of flexible operation on material response and subsequent damage. Moreover, a unified approach to creep-fatigue is significantly more effective at capturing this behavior.
There are several types of interactions that can drive additional damage. These include relatively well understood mechanisms, such as the effect of plasticity on primary creep and the effect of creep dwells on cyclic material properties. Other interactions that are less well understood include interruptions to the load during creep dwells and the effects of off-load periods on the overall creep rate.
This paper considers a constitutive approach to predict the modified creep rate due to load interruptions and off-load dwells using a backstress model. The backstress model is included in the calculation of inelastic strain rate equations, using a Chaboche type formulation. The model has been fitted to conventional material test data for typical superalloys used in gas turbine applications. To validate the approach, forward creep tests were conducted with varying interruptions to the load during the creep dwell period. These tests show a reduction in creep life and an increase in overall creep rate, when compared with the results for a constant stress and temperature condition. Previous work, presented by the authors , outlined a hypothesis that attributes the increase in overall creep rate to the influence of a recovery potential stress. This paper presents the subsequent work which demonstrates that the recovery potential stress can be defined by the difference between the applied stress and the backstress. It is shown that the dwell period between reload cycles is critical for calculating the recovery potential and overall creep rate.