Abstract
In the stress and fatigue evaluation of primary nuclear equipment, thermal stress calculation and fatigue curve have a great impact on the calculation results. There is a problem of stability in the step integral when calculating the thermal stress. The transient heat conduction changes rapidly near the initial temperature time point. It is critical to choose the right sub steps. In the nuclear code, the design fatigue curve in the air environment is given by applying a factor of 2 on stress or a factor of 20 on cycles. The factors are used to consider various influences such as data scatter, structure size, surface roughness and load. The factors of fatigue curves in ASME BPVC 2007 edition + 2009 addendum and French AFCEN’s RCC-M 2017 edition RPP 2 have changed. Due to the changes of the original test data and modificative factors, the new fatigue curve and the original fatigue curve are crossed. In code case N-761 of ASME, fatigue curves considering the influence of primary coolant environment are given under different strain rates.
In the calculation of thermal transient using commercial ANSYS software, the setting of time step has a great influence on the thermal stress. This paper compares and analyzes the influence of different calculation methods on the thermal stress at different positions of the model. The effect of different fatigue curves on fatigue life is also discussed considering different air design curve and different strain rate fatigue curve in water.
The results show that the fatigue results obtained by using the modified ASME or RCC-M air fatigue curve are smaller than those before the change when the usage factor is large to a certain extent. After considering the environmental fatigue curve, the fatigue results increase significantly. The accuracy of thermal stress calculation can be significantly improved by adding time points and refining time steps near the time points after the transient sharply changes, otherwise the influence of thermal stress on stress and fatigue results will be underestimated obviously.
