Comparative study on microstructure and hardness of P92 steel before and after long service

Due to its exceptional mechanical properties, the P92 steel is extensively utilized in power plants for high-temperature and high-pressure components, such as main steam and reheated steam pipes. This study aims to characterize the microstructure and Brinell hardness of domestic and imported P92 steel before and after 50 kh of service under identical conditions using SEM (scanning electron microscope), EBSD (electron back scattering diffraction), TEM (transmission electron microscope), and a Brinell hardness tester. The findings reveal that the hardness of domestic P92 steel decreases from 222 HB to 200 HB after 50,000 hours (50 kh) of service, while the imported P92 steel experiences a slight decrease from 205 HB to 203 HB. During prolonged service, the martensitic lath broadening rate in domestic P92 steel surpasses that in imported P92 steel, which primarily accounts for the significant decline in hardness. Additionally, the presence of large-sized Laves phases at matrix lath boundaries, lath bundle boundaries, and grain boundaries in both domestic and imported P92 steels also contributes to the reduction in hardness.


Introduction
As energy and environmental issues become increasingly prominent, there is an urgent need for power generation methods with higher energy conversion efficiency.In recent years, more and more ultrasupercritical units have been developed, significantly improving power generation efficiency [1,2].However, the high-temperature and high-pressure steam poses severe challenges to the service performance of materials.P92 steel is widely used in important components such as main steam and reheat steam pipelines in ultra-supercritical units due to its excellent high-temperature creep resistance and other comprehensive properties [3][4][5].Research shows that the heat-resistant martensitic steel P92 performs well due to its complex strengthening mechanisms, including solid solution strengthening, dislocation strengthening, precipitation phase dispersion strengthening, etc. Alloying elements in P92 steel, such as W, Mo, Cr, and Fe, form substitutional solid solutions, causing lattice distortion and strengthening the matrix.Nonmetallic small atoms like C, N, and B can form interstitial solid solutions with Fe.Elements like V and Nb act as strong carbon and nitrogen-forming elements, forming fine dispersed MX compounds that strengthen dispersion [6].Cr, Mo, Fe, and C can also form M 23 C 6 phases for precipitation strengthening [7].During prolonged high-temperature service, Laves phases may appear, typically precipitating at grain boundaries, effectively impeding dislocation movement and martensitic lath transformation, thus enhancing the stability of the microstructure through precipitation strengthening.Fine martensite and high-density dislocations prevent grain deformation and dislocation slip during high-temperature use, making them one of the most important strengthening mechanisms for P92 steel [8].P92 steel strengthens through fine lath-like martensite, grain boundaries, and precipitated phases.Studies on the microstructure evolution and strengthening mechanisms of P92 steel during hightemperature creep conducted by researchers have found that the growth of M 23 C 6 phases can decrease the creep fracture strength of P92 steel.It shows that the initial formation of Laves phases can effectively enhance the creep fracture strength.Still, once they grow to a certain size, they may reduce the creep fracture strength [9]. Lee et al. [10] have found that the formation of creep voids is related to the precipitation of Laves phases at grain boundaries, and the hardness of P92 steel decreases rapidly with the extensive precipitation of Laves phases.Research on domestic P92 steel subjected to 650°C aging treatment revealed that the coarsening rate of lave phases in domestic P92 steel is higher than in imported P92 steel, resulting in lower creep performance [11].
There is a considerable amount of research on the microstructure evolution and high-temperature creep performance of P92 steel after aging [12][13][14].However, most studies focus on imported P92 steel, with relatively few studies on domestic P92 steel.Moreover, many studies on P92 steel use long-term high-temperature creep tests to accelerate the simulation of microstructure transformation and performance changes during service.However, the duration of these tests is short, making it difficult to fully reflect the microstructure evolution and performance changes of P92 steel under long-term service conditions [15,16].This significantly constrains the effective development of heat-resistant steels for ultra-supercritical units in China.In response to this situation, this paper uses scanning electron microscopy (SEM) and transmission electron microscopy (TEM) to study the microstructure evolution of imported and domestic P92 steel before and after 50 kh of service.Hardness values are the only quantifiable representation of mechanical parameters achievable on-site, serving as an important indicator for evaluating whether materials meet standards and for metal supervision [17].This paper provides data support for promoting the domestication of P92 steel through intuitive Brinell hardness testing of the mechanical properties of domestic and imported P92 steel before and after service.

Material and experiment procedure
This experiment focuses on the left-side domestic P92 and right-side imported P92 main steam pipelines of an ultra-supercritical unit boiler in a domestic power plant.Northern Heavy Industries manufactures the domestic P92 steel, and the imported P92 steel is produced by the German company V&M.The study investigates the microstructure and Brinell hardness of the domestic and imported P92 steel after 50 kh of service.A comparison is made with the microstructure and Brinell hardness of the samples not in service (original samples retained during power plant construction).The designations for domestic and imported P92 steel, both before and after 50 kh of service, are as follows: Domestic P92 (0 h), Domestic P92 (50 kh), Imported P92 (0 h), and Imported P92 (50 kh).The actual operating temperature of the main steam pipeline is 605°C, and the actual working pressure is 27.56 MPa.The composition of the material before and after 50 kh of service is shown in Table 1.
Table 1.Chemical composition of P92 steel as received and after serviced for 50000 hours (wt %).Metallographic and hardness samples were cut from the cross-sections of domestic and imported P92 steel pipes using wire cutting.The samples were ground with sandpaper of different roughness (grit size) in a perpendicular cross direction until a 2500-grit finish was achieved.Subsequently, a polishing machine was used to polish the prepared samples to a mirror-like state.Before and after service, the samples were subjected to Brinell hardness testing using an XHBT-3000ZⅢ Brinell hardness machine to analyze the trend of hardness changes.For optical microscopy (OM) metallographic samples, ferric chloride hydrochloric acid solution was used for corrosion.Backscattered electron (BSE) samples for electron microscopy and electron backscatter diffraction (EBSD) samples were electrolytically polished with a 10% alcoholic chloric acid solution.The polishing voltage was set at 20 V, and the polishing time ranged from 15 to 20 seconds.The main focus was observing the size, content, and distribution of precipitated phases in the microstructure.Square thin slices with a thickness of 1 mm and a length of 10 mm were cut, mechanically polished to 60 μm, punched into circular slices with a diameter of 3 mm using a punch machine, and then further thinned using a double-jet thinning process to prepare transmission electron microscopy (TEM) samples.Through transmission electron microscopy, observations were made on the morphology and width of lath-like structures and the morphology and size of precipitated phases.

Microstructure analysis
Figure 1 shows the metallographic structures of domestic and imported P92 steel before and after 50 kh of service.Figures 1(a) and 1(c) depict the metallographic structures of domestic and imported P92 steel in the non-service state.It can be observed that the metallographic structures in the non-service state for both domestic and imported P92 steel consist of lath-like tempered martensite.The laths are densely packed, and no δ-ferrite is present.The lath boundaries are clear, and lath bundles are distinctly visible.Figures 1(b) and 1(d) show the metallographic structures of domestic and imported P92 steel after 50 kh of service.Both domestic and imported P92 steel maintain a well-preserved martensitic phase structure with no apparent degradation.Stable lath-like martensite is a primary strengthening mechanism for P92 steel.Martensite is formed by the shear transformation of austenite during rapid cooling.Multiple laths with similar orientations combine to form a lath bundle, and the size of these bundles is closely related to the size of individual laths.Previous research indicates that during the recovery process, precipitated phases, mainly M 23 C 6 , form at the original austenite grain boundaries, martensite lath boundaries, and subgrain boundaries, with the appearance of Laves phases and MX phases with increasing service time [13,18].

Hardness analysis
Figure 2 presents the Brinell hardness profiles of domestic and imported P92 steel before and after service.Before service, the hardness of domestic and imported P92 steel was approximately 222 HB and 205 HB, respectively.After 50 kh of service, the hardness values decrease to 200 HB for domestic P92 steel and 203 HB for imported P92 steel.The hardness reduction for domestic P92 steel is approximately 10%, while the hardness of imported P92 steel shows no significant decrease.

Backscattered electron analysis
Figure 3 comprises backscattered electron (BSE) images of domestic and imported P92 steel in different service states.BSE imaging brightness correlates with the atomic number of the sample, where higher atomic numbers result in increased brightness.This allows for qualitative analysis of precipitated phases, effectively distinguishing the locations and distributions of M 23 C 6 and Laves phases in P92 steel [19].Previous studies indicate that in P92 steel with non-service, the predominant precipitated phases are M 23 C 6 and MX, without Laves phases due to the tempering temperature being higher than the temperature for Laves phase formation (approximately 725°C) [20].Figures 3(a) and 3(b) show BSE images of domestic and imported P92 steel in the non-service state, revealing M 23 C 6 as the primary precipitated phase at martensite lath boundaries and interiors.The size of M 23 C 6 phases is similar for both domestic and imported P92 steel, approximately 110 μm.The phases are distributed mainly at lath boundaries, lath bundle boundaries, and original austenite grain boundaries.The presence of abundant dislocation structures at these interfaces provides significant energy fluctuations, facilitating the aggregation of alloying elements at boundaries and promoting the generation of precipitated phases.According to the chemical composition of P92 steel, MX phases (V, Nb) (C, N) should form; however, these may be challenging to detect under scanning electron microscopy conditions due to their low content and small size.Figures 3(c) and 3(d) show BSE images of domestic and imported P92 steel after 50 kh of service.Laves phases, characterized by a chemical composition mainly of Fe 2 W, exhibit higher brightness than M 23 C 6, characterized by a chemical composition mainly of Cr 23 C 6 in BSE images [21].The growth mechanism of Laves phases at high-temperatures primarily involves grain boundary diffusion [16], leading to a faster growth rate than M 23 C 6 .Figure 3 reveals that the size of Laves phases in imported P92 steel is slightly larger than in domestic P92 steel.Compared to the non-service P92 steel, Laves phases have grown significantly, with sizes surpassing 200 nm.This exceeds the beneficial strengthening limit size for Laves phases mentioned by Lee et al. [10] (130 nm critical value).The presence of large-sized Laves phases formed in the matrix after long-term service is identified as one of the contributing factors to the hardness reduction in both domestic and imported P92 steel.

Transmission electron microscopy analysis
Figure 4 presents transmission electron microscopy (TEM) images of domestic and imported P92 steel in different service states, and the corresponding statistics of lath width variations are shown in Figure 5. Figures 4(a) and 4(c) illustrate that martensite laths of domestic and imported P92 steel are straight in the non-service state, with minimal evidence of precipitated phases.Figures 4(b) and 4(d) show TEM images after 50 kh of service, revealing the presence of precipitated phases at lath boundaries.The quantity and size of precipitated phases in domestic P92 steel are noticeably higher than in imported P92 steel.Image Pro software was used to analyze the width changes of martensite laths statistically, and the results in Figure 5 indicate that the average width of martensite laths for domestic P92 steel increased from approximately 250 nm before service to about 520 nm after 50 kh of service-a widening rate exceeding 100%.The average width of martensite laths for imported P92 steel increased from around 260 nm before service to about 440 nm after 50 kh of service-a widening rate below 70%.These results show that the stability of martensite lath bundles in imported P92 steel is higher than in domestic P92 steel.The fine structure of martensitic laths is one of the most crucial strengthening mechanisms for P92 steel.P92 steel undergoes a heat treatment process involving quenching and tempering, forming an oversaturated solid solution martensite lath structure.Combined with abundant lath boundaries, dislocation structures, and alloying elements dissolved within, this structure imparts high strength and hardness to P92 steel.As P92 steel operates at high-temperatures, the martensite lath structure widens.Previous research indicates that martensite lath strengthening accounts for over 60% of the strengthening mechanisms in P92 steel, and the widening of martensite laths is a primary reason for the decrease in strength and hardness of P92 steel [8].

Electron back scattered diffraction analysis
Figure 6 shows the EBSD grain boundary maps of domestically produced and imported P92 steel before and after 50 kh of service.The red lines represent small-angle grain boundaries with orientation differences between 2° and 10°, while the black lines represent large-angle grain boundaries with orientation differences greater than 10°.Figures 6(a) and 6(b) depict the EBSD grain boundary orientation maps for domestic P92 steel before and after service.The black large-angle grain boundaries correspond to the grain boundaries of the original austenite or lath bundle interfaces.In contrast, the red small-angle grain boundaries correspond to lath block interfaces.Small-angle grain boundaries are unstable subgrain boundaries, and their changes over service time are more significant compared to large-angle grain boundaries.Statistical results using Channel 5 software in Figure 6 indicate that the proportion of small-angle grain boundaries in domestic P92 steel increased from approximately 47% before service to about 53% after service, an increase of 6%.In imported P92 steel, the proportion of small-angle grain boundaries increased from 49% before service to 52% after service, an increase of 3%.When the laths are narrow, EBSD is challenging to measure the orientation difference between adjacent laths, yielding information only on the orientation difference of lath blocks.As the laths widen, EBSD gradually measures the orientation difference between adjacent laths, manifested as an increase in the number of small-angle grain boundaries in EBSD.Small-angle grain boundaries are a concentrated manifestation of interfaces such as lath block or lath boundaries in P92 steel.The more severe the widening of the laths is, the greater the number of small-angle grain boundaries is.After long-term hightemperature service, the martensite structure recovers, increasing the measurable subgrain boundaries such as lath bundle or lath boundaries, reflected in the number of small-angle grain boundaries.The changes in small-angle grain boundaries in domestic and imported P92 steel are relatively small.However, regarding numerical changes, the variation in small-angle grain boundaries in imported P92 steel is smaller, indicating a more stable lath structure.

Figure 2 .
Figure 2. Brinell hardness graph of domestic and imported P92 steel under different service states.

Figure 5 .
Figure 5.The width variation of lath martensite of domestic and imported P92 steel under different service states.