Abstract
Many power plant components, such as steam generators, are exposed to creep regime temperatures during operation. Creep deformation is a major life limiting factor for such components. Therefore, estimation of creep life is important for ensuring safe operation of power plants. Austenitic steel is commonly used for steam generator components which also include thick sections of multi-pass welds. Creep deformation and damage is a biggerproblem in multi-pass welds due to the material heterogeneity of such welds along with presence of defects and weld residual stresses. Investigation of creep damage is particularly complex in such welds. A robust understanding of creep deformation is the first step to any creep damage assessment. A realistic model of creep deformation of multi-pass weldments is complicated by the presence of different material zones, namely, base metal, weld and the heat affected zone (HAZ). This thesis investigates the existence and extent of elastic-plastic and creep properties variation across a 316H steel multi-pass weld and also explores effects of such mismatch on the overall creep response of the weldment. The scope of the work is limited to the creep deformation
behaviour of 316H multi-pass welds and therefore tertiary creep was not explored in this treatise.
Estimates for stress and strain concentration effects were obtained using simplified finite element (FE) models and the results were compared against weld stress concentration factors from earlier work and engineering assessment codes. Localised creep tests were conducted on small test specimens extracted from different locations of an ex-service 316H weld HAZ section and thereby minimum creep rate in the HAZ was characterised as a function of the distance
from the weld fusion line. DIC creep tests were then conducted on crossweld specimens of the aforementioned weld at different stress levels and using benchmarked finite element models, the full field strain and stress distributions in the specimens were studied. Also, DIC creep testing and finite element modelling were used to investigate any augmented effects of weld cap geometry on stress concentrations due to existing material mismatch across the weld fusion line of a 316H multi-pass weld.
Results from the analytical models showed that upper bound estimates of stress concentration factors for the worst level of creep rate mismatch across a weld interface will always be lower than the stress concentration factors used for dissimilar welds. A method of creep rupture design of weldments was proposed which considers both the rupture stress in the creep strong material
as well as creep strain rates in the creep weak material near the weld interface. DIC creep tests of crossweld samples of 316H weld depicted the extent and deformation behaviour of a creep strong material zone in the HAZ section. Finite element model results showed good correspondence with the tests and it was demonstrated that a combination of FE modelling and DIC tests can reveal true stress as well as strain field information in a test specimen. DIC creep tests and analytical modelling of a crossweld specimen with a simulated weld cap showed that effect of creep mismatch remains insignificant and the stress concentration due to only the weld cap geometry remains unchanged from that of a homogeneous metal specimen.
The work showed that although there can be significant difference in creep deformation behaviour between the base metal and the weld metal in a 316H multi-pass weld, its effects on stress and strain rate distributions near the weld fusion line and in the overall weld are insignificant. Also, the work demonstrated the use of a combination of DIC creep tests and FE modelling in studying weld creep and determination of the extent of a creep strong region in
the weld where creep rates gradually vary. Stress concentration factors due to weld caps will be dominated by the geometry effects with insignificant contribution from the creep rate mismatch between the weld and the HAZ.
| Date of Award | 23 Jun 2020 |
|---|---|
| Original language | English |
| Awarding Institution |
|
| Supervisor | Christopher E Truman (Supervisor) & David M Knowles (Supervisor) |