Microstructural evolution in boron-containing high-chromium steels for power plant applications

Martensitic 9-12% Cr steels have been the key materials used to increase the efficiency of ultra-super-critical (USC) power plants, offering a good combination of high steam oxidation resistance and high creep strength. Martensitic 9-12% Cr steels have been the key materials used to increase the efficiency of ultra-super-critical (USC) power plants, offering a good combination of high steam oxidation resistance and high creep strength. The addition of boron was found to stabilize the martensitic laths by decreasing the coarsening rate of M23C6 carbides, which act as pinning points in the microstructure and lead to an improved creep behaviour in these new enhanced martensitic steels.

The aim of this research project was to obtain a deep and quantified understanding of microstructural changes in boron-containing high Cr ferritic-martensitic power plant steels, as a function of pre-service heat treatment, stress, time and temperature. The creep strength, which is the main design criteria for this class of alloys, depends on the stability of the microstructure, which consists of tempered martensite and a fine dispersion of carbide precipitates. During this study a comprehensive microstructural characterisation on the virgin and creep-tested samples was undertaken in order to provide valid and reliable data to correlate the microstructural changes with creep behavior. This can provide steel manufacturers with useful information to optimise the manufacturing route for fabrication.

Advanced characterisation techniques have been employed to investigate the effect of normalizing and tempering on the microstructure of a MARBN steel, and the stability of second phase particles after pre-service heat treatment has been investigated. It has been observed that the normalizing temperature has a significant impact on the size of prior austenite grain boundaries, which increase significantly at higher normalizing temperature. Early precipitation of Laves phase has been observed after tempering. The processing conditions were observed to have a significant impact on the area percentage of this phase. It has been identified that normalizing temperature doesn’t seem to significantly affect the precipitation of Laves, whereas the tempering temperature and time were found to influence the precipitation of Fe₂W precipitates. Furthermore, the tempering temperature was observed to have a direct impact on the number of M₂₃C₆ and MX carbonitrides precipitated during tempering. In fact, a reduction of M₂₃C₆ particles and an increase of MX carbonitrades has been observed by increasing the tempering temperature.

The effect of manufacturing process on the microstructural evolution and creep life of two modified FB2 steels, with the same chemical composition but different manufacturing process, was studied using scanning and transmission electron microscopy (SEM and TEM), focused ion beam microscopy (FIB), energy dispersive x-ray spectroscopy (EDS), electron backscattered diffraction (EBSD), X-ray diffraction (XRD) and thermodynamic equilibrium calculations. Investigations have proven that the additional melting stage before the pre heattreatment underwent by one of the two steels has significantly affected the boron nitride (BN) characteristics in the material. In fact, it was observed that the amount of BN particles is much higher in the steel that hasn’t experienced the re-melting process (sample C). In turn, the faster evolution rate of the M₂₃C₆ particles in the C sample during the exposure to the creep conditions correlates very well with the amount of BN particles observed in the original conditions. It is believed that more boron is trapped in the BN phase in the C sample, leaving less boron in the solid solution to stabilize the M₂₃C₆ particles from coarsening and in turn influence the population of creep resisting VN particles.

Analysis of the gauge sections using a combined in-lens SEM/EDS approach showed that cavities containing BN particles were also present within the C steel, while no cavities were detected in the P steel, which is the steel that has experience the re-melting stage. Hence, it is believed that the significantly shorter creep life of the C steel compared to the P sample is determinated by a combined effect of the higher BN particle number density and the formation of cavities containing BN particles within the gauge area, which contribute to the real fracture. The lower BN content in the P sample, as well as the absence of any cavities, instead, could result in an increased concentration of soluble boron available to stabilize the M₂₃C₆ particles from coarsening during creep, as well as more nitrogen, which allowed a more stable MX distribution, resulting in the observed increased creep life.

Advanced characterisation techniques have also been employed to understand the reasons why a modified St13 steel failed the ultrasonic test performed by the manufacturer for a first quality check before pre-heat treatment. Bright field reflected light microscopy revealed a very heterogeneous microstructure with large grains (sized up to 4 mm) mixed with area with much smaller grains. A combination of selected area diffraction and STEM-EDS analysis was used to identify the large VN “decorating” the grain boundaries, which have been identified as the possible reasons of the ultrasonic test failing.

Ni enriched regions were also detected in the modified St13 steels investigated, and TEM diffraction confirmed their fcc crystal structure to be consistent with γ austenite. An orientation relationship, corresponding to Kurdjumov-Sachs (K-S), has also been identified between the bcc martensitic matrix and the fcc retained austenite. The study of the evolution of Ni-rich regions has also shown that the exposure to the creep conditions significantly accellerates the coarsening of the Ni enriched regions in the modified St13 steel investigated.