| Summary: | Components in advanced gas cooled reactor (AGR) operating at elevated temperatures in the range of 500-650°C are typically susceptible to the initiation and growth of cracks due to creep. Type 316H stainless steel steam headers after a long term service are susceptible to reheat cracking in the vicinity of the weld driven by the presence of welding residual stresses. For this reason, this research has focused on developing pragmatic numerical methods for predicting creep crack growth behaviour of welded components containing residual stresses, using a simplified continuum damage and fracture mechanics method. The work presented had three main aims. The first was to derive a comprehensive set of plastic η factors for standard fracture mechanics geometries containing welds. The impetus for this was to improve material crack growth characterisation for welds by improving the creep C* solutions for these geometries that are presently recommended in standard codes of practice for creep crack growth testing. The second part was to experimentally examine appropriate creep material properties for as-received and service-exposed 316H stainless steels containing welds for use in numerical modelling and predictive methods for creep crack growth in a real component. The third was to develop and validate a simplified method of simulating residual stresses and creep crack growth behaviour in an ex-service AISI 316H weld header with reheat cracking. This approach simulates the presence of residual stresses using appropriate loading and boundary conditions in actual components that undergo reheat cracking without the need to develop full weld simulations to quantify them. The creep crack growth behaviour was studied using two methods based on the theories of fracture mechanics and continuum damage mechanics. Fracture mechanics parameter C* was firstly used to examine the approximate crack growth rate using the reference stress approach and approximate NSW model. The second method was to predict long term cracking by using a simplified continuum damage mechanics model, with a consideration of stress relaxation. For this purpose, a simplified multi-axial ductility exhaustion model was developed and implemented in an Abaqus user subroutine, taking into account the changes in the ex-service creep properties and the effect of reduction in creep ductility under low loads and long term operation at service temperatures. Resulting from the findings, the task was to identify the geometric and the material reasons of how and why the crack growth follows a path of least resistance and higher constraint which did not necessarily mean growing through the welds or the heat affected zone region.
|
|---|
