Components used in the power generation sector are often continuously exposed to high temperatures and corrosive environments. Failure processes, such as net section rupture, creep crack growth or fatigue crack growth therefore occur within the high temperature regime. The presence of residual stresses plays an important role in the subsequent failure of engineering components and structures. Residual stresses can arise from almost all manufacturing and fabrication processes and can also arise during service. Tensile residual stresses may combine with in-service loads to promote failure at a load the designer would view as safe. A quantitative understanding of how residual stresses interact with applied service loads is thus required for accurate safety assessments.
In this paper a test rig based on a three bar structural model is used to introduce long range residual stresses in a 316H steel C(T) specimen at high temperature. The residual stresses induced are characterized easily without use of time consuming residual stress measurement techniques. The complete test rig is then subjected to an applied load. The magnitude of the residual and applied stress in the 316H C(T) specimen is a function of the initial misfit displacement, applied load and relative stiffness of the components of the test rig. The experimental results show that a test rig with a higher elastic follow-up value will have more crack growth compared to a rig with a lower elastic follow-up. Also, both tests demonstrate that as the crack grows, relaxation of residual stress in the C(T) specimen occurs, and it is compensated by a change in residual stress distribution in other parts of the rig. Furthermore, creep crack initiation data is compared with load controlled tests conducted. It is found that the time for the crack to initiate is increased in the case of mixed boundary conditions compared to load controlled conditions.