Microstructural evolution of a 12CrNiMoV power plant steel

To increase the knowledge of the microstructural evolution of a 12CrNiMoV power plant steel under elevated temperature and stress conditions; the results of microstructural studies carried out on creep rupture samples provided by GE Power, and on furnace controlled isothermally heat treated samples generated at Loughborough University are presented. Multiple investigative avenues are presented to cover the range of sample material test and heat treatment conditions. A variety of techniques are used including Scanning (SEM), Transmission (TEM) Electron Microscopy, Focused Ion Beam (FIB) microscopy and Neutron diffraction. Developments in FIB microscopy through the additional use of an insulated enhanced etching are presented allowing separation of intermetallic and carbide phases for imaging and quantification purposes.

The impact of a variation in tempering temperature and time between two suppliers both meeting the minimum heat treatment requirements and the minimum mechanical release properties provided the first avenue of study. The subtle differences identified in the microstructure, and the precipitate phase fractions and compositions showed no impact on short term creep properties and that a double tempering treatment, often used for this alloy, is not necessarily required to produce a fully martensitic structure while avoiding retained austenite.

A forensic investigation is undertaken to differentiate between three creep rupture samples that demonstrate greater than expected scatter in rupture strength. One sample is shown not to be representative of the material heat and demonstrates no microstructure features that could be identified as the cause of failure. Microstructural evidence for the scatter shown has been suggested that could be used for alloy development to enhance creep strength by more than 15%.

Thermodynamic equilibrium calculations are presented using heat specific chemical compositions along with an initial attempt at precipitate age modelling to study the effects of the heat-to-heat variations in alloying chemistry on phase stability. The predicted effects of formation of secondary phases such as M₂₃C₆, Laves phase and MX are described in context of chemical composition variation, heat treatment variation and the impact on microstructural transformation temperatures. The predictions showed some consistency with subsequent experimental results obtained from the current research.

Using the creep exposed samples and isothermally aged samples, comparisons between the evolution behaviour of the phases central to the alloys long-term creep strength such as M₂₃C₆, Laves, MX and M₂X with respect to stress, time and temperature are presented. Clear differences are shown between the stressed gauge sections and the theoretically un-stressed head /grip sections of the creep rupture samples with respect to both matrix microstructure and secondary precipitates. Greater microstructural recovery is evident in the form of greater subgrain formation alongside a reduction in grain average misorientation, whilst the kinetics of both M₂₃C₆ and Laves phase are shown to be sensitive to the stress component during creep ageing. A potential interrelational existence between Laves and M₂₃C₆ is demonstrated, however, it is not yet understood if a loss of carbides rich in constituent elements of Laves phase drives Laves phase evolution, or if Laves phase feeds from M₂₃C₆ in a dominant fashion. The control of coarsening of Laves phase may directly impact the effective microstructural control of M₂₃C₆ and other Mo or Cr containing precipitates and subsequently the long term creep strength of the alloy.

Changes in the martensitic matrix are shown to be limited during elevated temperature isothermal aging whilst the exposure to elevated temperature and stress, leading to creep, is shown to result in more significant matrix recovery. Clear microstructural recovery is evidenced by the loss of hardness with respect to time and temperature but indication of matrix recovery was not seen quantitatively until the assessment of GB statistics and misorientation data was explored. The resulting relationship between the hardness, matrix recovery and M₂₃C₆ precipitation distribution statistics is presented.

During chemical element mapping of the creep samples, unexpected segregation of Ni was discovered. Ni specific element mapping found regions of enrichment with up to 10% Ni existed in a chemically distinct phase from the surrounding microstructure. The identification and characterisation of this Fe-Cr-Ni phase are presented along with the preferred formation conditions following experimental assessment of the isothermally aged samples. Potential phase impacts are proposed based on thermodynamic calculation and physical property testing.

The increased knowledge of the evolution mechanisms of this alloy presented in this thesis can be used for alloy development, to improve precipitation models and improve alloy specific understanding of the mechanisms for the loss of creep strength during service. This research project has highlighted the potential impacts of surpassing the current maximum alloy usage temperature and demonstrated potential development for greater microstructural and subsequent mechanical control of the alloy that could allow for increased elevated temperature creep properties.