Mechanical Properties of P91 Steel (X10CrMoVNb9-1) during Simulated Operation in a Hydrogen-Containing Environment

P91 steel (X10CrMoVNb9-1) is widely used in the energy industry. It is characterized by good mechanical properties, creep resistance, corrosion resistance, impact toughness, and resistance to thermal fatigue. Due to their operating conditions and martensitic structure, components made from P91 steel are often subject to damage related to the presence of hydrogen. This article compares the results of the mechanical properties evaluation for P91 steel in an aggressive solution charged under load and without load. Based on the research, it was found that the hydrogen environment significantly affects the mechanical properties of P91 steel, reducing strength and yield strength, and decreasing ductility. It was revealed that in samples tested after 72 h without preloading, the tensile strength decreased by 1.5%, and the elongation decreased by about 29% for the sample, compared to the delivered condition sample. Under loaded conditions, the difference in tensile strength increased by approximately 8%, while elongation increased by nearly 50%

In the as-delivered condition, P91 steel is normalized at 1040-1080 • C, which results in the dissolution of carbides without significant grain growth.Then, at 750-780 • C, carbides precipitate uniformly throughout the martensite tempering process [1,[9][10][11].These are M23C6 carbide types rich in V/Nb.The microstructure also contains rich carbonitrides of the MX type (M = V or Cr; X = C or N).The presence of these precipitates significantly increases creep rupture strength through precipitation hardening.In the as-delivered condition, the strength of P91 steel is a minimum of 585 MPa, and the yield strength is no less than 415 MPa, with an A5 elongation of 20% [12][13][14][15][16].
Due to their operating conditions and martensitic structure, components made from P91 steel are also subject to damage related to the presence of hydrogen.Hydrogen damage is a very unfavorable phenomenon affecting many industries [17][18][19].
Factors determining the phenomena associated with the presence of hydrogen include hydrogen penetration into metals, hydrogen diffusion, hydrogen trapping, and stress and strain [20].
The presence of hydrogen in a material can cause hydrogen gas embrittlement, decreased ductility, internal cracking and delamination, segregation cracking, snowflakes and fish eyes, cold cracking, e.g., following a welding process [21][22][23], stress cracking accompanying the formation of hydrides, and the action of hydrogen during transformations, such as martensitic transformation, as well as SOHIC (Stress-Oriented Hydrogen-Induced Cracking) and HIC (Hydrogen-Induced Cracking) [24].
The literature also presents numerous mechanisms describing various phenomena related to the influence of hydrogen within materials.Among the commonly accepted mechanisms are three primary ones, which can act either independently or in concert.These are hydrogen-enhanced local plasticity (HELP), hydrogen-enhanced decohesion (HEDE), and adsorption-induced dislocation emission (AIDE) [25].
Hydrogen-enhanced local plasticity (HELP) is attributed to increased dislocation mobility in the presence of hydrogen.The accumulation of hydrogen reduces the dislocation stress field, thereby locally decreasing the dislocation cross-stress [26].Consequently, a lower external stress is required to induce dislocation movement compared to an environment without hydrogen.
Another phenomenon contributing to embrittlement is hydrogen-enhanced decohesion (HEDE).HEDE describes the embrittlement of metals due to reduced bonding forces between matrix atoms caused by local accumulations of stored hydrogen.This local accumulation, often occurring at grain boundaries or phase interfaces, significantly lowers the energy required to initiate crack propagation [27].
In the case of adsorption-induced dislocation (AIDE), hydrogen embrittlement arises from a reduction in surface energy caused by hydrogen adsorption.The accumulation of hydrogen around precipitates or pores significantly reduces surface energy, thereby promoting microcrack formation.[28] Other descriptions of hydrogen-related mechanisms are provided in the work [15].
According to [19], the hydrogen damage process can be divided into hydrogen absorption, diffusion, and trapping, which eventually leads to cracking.There are several methods for evaluating a material's resistance to hydrogen-induced damage: testing under constant stress on smooth or notched specimens, testing under constant strain, e.g., under constant tensile strain or at a given bending rate, and testing under slow load increase [24].An important element in the determination of the hydrogen resistance of a material are corrosion tests: SSC (Sulphide Stress Cracking), according to NACE TM 0177-96, and HIC, according to NACE TM 208 [29][30][31].
However, SSC and HIC tests are not static tests, which means they do not provide an answer to the behavior of the material during simultaneous loading.An interesting test is the Slow Strain Rate Test (SSRT) in solution.There are numerous articles in the literature on SSRT tests conducted on materials dedicated to the energy industry, e.g., on hydrogen gas transportation pipelines [32][33][34][35] with an austenitic or ferritic-pearlitic structure, but there are definitely fewer studies on martensitic steels.
Interesting research on martensitic steels is described in the paper in [36].The authors carried out the SSRT test with 3.5% NaCl, 1 g/L thiourea, and 1 mA/cm 2 current density on two low-carbon martensitic steels Fe-0.068C-0.39 Ti (Ti-containing steel) and Fe-0.069C-0.45 Mo (Mo-containing steel).They found that the crack initiation in Mo-containing steel originated from Mo 2 C carbides, accompanied by decohesion at the interface between the martensitic matrix and Mo 2 C carbides, and that the hydrogen trapping capacity at the designated hydrogen trap sites was a determinant factor for H-assisted crack initiation and propagation in martensitic steels.They also found that the Ti-containing steel indicates a mitigated hydrogen embrittlement susceptibility, controlled by carbon vacancies inside TiC acting as strong hydrogen trap sites.
Other interesting research on the influence of the strain rate was presented in the article in [37].According to this survey, the yield strength was not affected considerably by temperature (the experiment was conducted at 25 • C and 50 • C) and only affected by strain rate due to the effect on dislocation annihilation/motion, and the hydrogen embrittlement susceptibility increased with decreasing strain rate.They also found that more pronounced hydrogen embrittlement in low strain rates is due to the ability of hydrogen to interact with mobile dislocations, which leads to hydrogen-induced fracture.
Equally satisfactory research on the SSRT test with different current densities from 0.1 to 10 mA/cm 2 was carried out on two martensitic steels with different Mo content (0.43 and 1.06 wt.%), which are presented in the paper in [38].The authors showed that the crack growth rate increased and the values of stress intensity factors K IH and K Imax decreased with the increase in pre-charged hydrogen concentration.The steel exhibiting a higher molybdenum content demonstrated a significantly reduced crack growth rate compared to the steel with a lower Mo content.They showed it could be attributed to the presence of a greater density of nano-sized precipitates.These precipitates act as effective hydrogen trapping sites, effectively mitigating the detrimental effects of hydrogen on the crack propagation rate.
The only available study on the behavior of P91 steel during the SSRT test is described in [39], where Bharasi et al. conducted tests in a 1-4 M sodium hydroxide (NaOH) solution.The authors observed secondary stress corrosion cracks in all specimens tested in 1-4 M NaOH and the number of secondary cracks increased with an increase in concentration up to 3 M. Tensile test data showed that ductility decreased with an increasing concentration of NaOH up to 3 M, and practically remained the same for 4 M NaOH solution.
Despite numerous corrosion tests, the behavior of a material exposed to hydrogen is sometimes difficult to predict.Similarly, due to the complexity of factors affecting the acceleration or slowing down of structural degradation, it is very difficult to correctly estimate the effect of hydrogen on the durability of an industrial component or facility.In order to select the best possible test methodology, it is important to identify the exact operating conditions (including the operating pressure, temperature, and environment).
The results of available research show that there are few articles about the corrosion resistance of P91 steel at room temperature and elevated temperatures, however, there is very little information on the research around martensite P91 steel under the influence of hydrogen under load.
It therefore seems necessary to conduct research that will contribute to the description of the microstructure of P91 steel in the SSRT test.This paper compares the results of tests aimed at determining the mechanical properties of P91 steel following its charging with hydrogen, both under load and without load.The results unambiguously confirm a deterioration in the mechanical properties of P91 steel under the influence of hydrogen, in particular under load.

Materials and Methods
High-alloyed steel P91 (X10CrMoVNb9-1) with a martensitic structure designed for service at elevated temperatures was used for the tests.The structure in the as-delivered condition is shown in Figure 2. The tests were carried out on cylindrical specimens (in accordance with PN-EN ISO 6892-1) [40] with the geometry shown in Figure 3.

Materials and Methods
High-alloyed steel P91 (X10CrMoVNb9-1) with a martensitic structure desig service at elevated temperatures was used for the tests.The structure in the as-d condition is shown in Figure 2. The tests were carried out on cylindrical specim accordance with PN-EN ISO 6892-1) [40] with the geometry shown in Figure 3.The tests were carried out on a specially prepared stand (Figure 4) with the M fatigue machine (250 kN) (MTS Systems, Eden Prairie, MN, USA).The stand allo hydrogen charging of the specimens while applying a tensile preload of a specifie The tests were carried out at the Faculty of Materials Engineering of the Silesian sity of Technology on a stand based on the MTS-810 250 kN fatigue machine.Th enables electrolytic rehydration in a solution of 0.5 M H2SO4 and 25 mg/L As2O3.T rent flow used in the tests was 10 mA/cm 2 .These conditions simulate the oper pipelines in power installations and tanks for transporting hydrogen under high p

Materials and Methods
High-alloyed steel P91 (X10CrMoVNb9-1) with a martensitic structure desig service at elevated temperatures was used for the tests.The structure in the as-d condition is shown in Figure 2. The tests were carried out on cylindrical specim accordance with PN-EN ISO 6892-1) [40] with the geometry shown in Figure 3.The tests were carried out on a specially prepared stand (Figure 4) with the M fatigue machine (250 kN) (MTS Systems, Eden Prairie, MN, USA).The stand allo hydrogen charging of the specimens while applying a tensile preload of a specifie The tests were carried out at the Faculty of Materials Engineering of the Silesian sity of Technology on a stand based on the MTS-810 250 kN fatigue machine.Th enables electrolytic rehydration in a solution of 0.5 M H2SO4 and 25 mg/L As2O3.T rent flow used in the tests was 10 mA/cm 2 .These conditions simulate the oper pipelines in power installations and tanks for transporting hydrogen under high p The tests were carried out on a specially prepared stand (Figure 4) with the MTS-810 fatigue machine (250 kN) (MTS Systems, Eden Prairie, MN, USA).The stand allowed the hydrogen charging of the specimens while applying a tensile preload of a specified value.The tests were carried out at the Faculty of Materials Engineering of the Silesian University of Technology on a stand based on the MTS-810 250 kN fatigue machine.The stand enables electrolytic rehydration in a solution of 0.5 M H 2 SO 4 and 25 mg/L As 2 O 3 .The current flow used in the tests was 10 mA/cm 2 .These conditions simulate the operation of pipelines in power installations and tanks for transporting hydrogen under high pressure.
In the first part, a static tensile test was carried out on the material in the as-delivered condition, in accordance with PN-EN ISO 6892-1, to determine the basic mechanical properties of the P91 steel tested.On this basis, the value of the preload to be applied during the hydrogen charging of other specimens was determined.The preload value selected was 70% of yield strength Rp 0.2 , which corresponded to approx.380 MPa.
The remaining specimens (18) were electrolytically charged in 0.5 M H 2 O + 25 mg/L As 2 O 3 solution.Half of the specimens (9) were charged under load, and the other half (9) without load.In both groups, 3 specimens were hydrogen-charged for 24 h, 3 specimens for 48 h, and 3 specimens for 72 h.The results presented below show the average for each category of specimens.
Following the hydrogen charging process, static tensile strength tests were carried out on the specimens under displacement control at a constant rate of 4 mm/min.
Throughout the tests, the following values were continuously recorded: time, displacement, force, and strain.An MTS-623.11c.20 extensometer (MTS Systems, Eden Prairie, MN, USA)with a gauge length of 25 mm was used to measure strain values.
Based on the results obtained, static tensile stress diagrams were developed, also including accurate measurements of material elongation following the hydrogen charging with and without load.
On this basis, the basic strength and plasticity properties of P91 steel were determined.Subsequently, the fracture surfaces and transverse and longitudinal cross-sections were subjected to metallographic examinations under a Hitachi S-3400N scanning microscope (Hitachi, Ltd., Tokyo, Japan) using the secondary electron detection technique.Images were recorded in the secondary electron mode and at a voltage accelerating the electron beam to 25 keV.In the first part, a static tensile test was carried out on the material in the as-delivered condition, in accordance with PN-EN ISO 6892-1, to determine the basic mechanical properties of the P91 steel tested.On this basis, the value of the preload to be applied during the hydrogen charging of other specimens was determined.The preload value selected was 70% of yield strength Rp0.2, which corresponded to approx.380 MPa.
The remaining specimens (18) were electrolytically charged in 0.5 M H2O + 25 mg/L As2O3 solution.Half of the specimens (9) were charged under load, and the other half (9) without load.In both groups, 3 specimens were hydrogen-charged for 24 h, 3 specimens for 48 h, and 3 specimens for 72 h.The results presented below show the average for each category of specimens.
Following the hydrogen charging process, static tensile strength tests were carried out on the specimens under displacement control at a constant rate of 4 mm/min.
Throughout the tests, the following values were continuously recorded: time, displacement, force, and strain.An MTS-623.11c.20 extensometer (MTS Systems, Eden Prairie, MN, USA)with a gauge length of 25 mm was used to measure strain values.
Based on the results obtained, static tensile stress diagrams were developed, also including accurate measurements of material elongation following the hydrogen charging with and without load.
On this basis, the basic strength and plasticity properties of P91 steel were determined.Subsequently, the fracture surfaces and transverse and longitudinal cross-sections were subjected to metallographic examinations under a Hitachi S-3400N scanning microscope (Hitachi, Ltd., Tokyo, Japan) using the secondary electron detection technique.Images were recorded in the secondary electron mode and at a voltage accelerating the electron beam to 25 keV.

Results and Discussion
The study aimed to assess the mechanical properties and determine the structural

Results and Discussion
The study aimed to assess the mechanical properties and determine the structural phenomena that determine the hydrogen embrittlement of martensitic P91 steel (Figure 3, 14a) exposed to a strongly hydrogenating environment.Such conditions often occur in power installations, and, due to its high strength properties, it can be used in installations for transporting compressed hydrogen.For this reason, it is important to understand and confirm the mechanisms that affect the tendency toward hydrogen-assisted cracking.
The results of the static tensile strength tests on P91 steel in the as-delivered condition enabled the determination of the basic mechanical properties of the material tested.The values obtained were later used in further testing phases.Table 1 and Figure 5 compare the results obtained for the specimens tested in the as-delivered condition, and those tested following hydrogen charging without load.Table 2 and Figure 6 compare the results obtained for the specimens tested in delivered condition and those tested following hydrogen charging under load.Table 2 and Figure 6 compare the results obtained for the specimens tested in the as-delivered condition and those tested following hydrogen charging under load.The analysis of the results obtained during the tests (for Rm, Re, A5, a indicates that hydrogen causes a reduction in both strength and plasticity-re ties (Figures 7-10).A slight reduction in the material's strength was identified h of exposure to hydrogen.In the case of the tests without preloading, the s The analysis of the results obtained during the tests (for R m , R e , A 5 , and Z) clearly indicates that hydrogen causes a reduction in both strength and plasticity-related properties (Figures 7-10).A slight reduction in the material's strength was identified even after 24 h of exposure to hydrogen.In the case of the tests without preloading, the strength of the material decreased as the charging time increased and fell to 672 MPa after 72 h, which means a decrease of 10 MPa (Figure 7).It should be concluded that, under these conditions, with strength as the criterion, P91 steel can be deemed resistant to hydrogen-containing environments.Much worse strength results were obtained for the specimens that were hydrogen-charged under constant load and then subjected to the tensile strength tests.
vironments.Much worse strength results were obtained for the specimens that were gen-charged under constant load and then subjected to the tensile strength tests.
As early as after 24 h, the material's strength decreased to 654 MPa, and after decreased to 627 MPa.This confirms that hydrogen affects the mechanical prope P91 steel, especially during service and operation.
As for the yield strength (Rp0.2), which is 546 MPa in the as-delivered condi decreased as the charging time increased, falling to 501 MPa after 72 h of charging w load (Figure 8).The analysis of the results obtained for the specimens charged und showed a reduction in yield strength compared to the value obtained for the mat the as-delivered condition (by 20 MPa), but an increase in yield strength compared specimens charged without load.The difference in Rp0.2 between the specimens c without load and under load was 13.8 MPa for the charging time of 24 h, and 26.1 M the charging time of 72 h (Figure 8).The evaluation of plastic properties (A, Z) showed a significant decrease in t tility of the specimens exposed to hydrogen in comparison with the specimens in delivered condition (Figures 8 and 9).In particular, a considerable deterioration in properties was observed for the specimens that were hydrogen-charged under lo specimens charged without load for 72 h showed a 6% decrease in percentage elon A, which is 29% less compared to the as-delivered condition (Figure 9).The elong was 11%, which was a 50% decrease compared to the as-delivered condition.delivered condition (Figures 8 and 9).In particular, a considerable deterioration in properties was observed for the specimens that were hydrogen-charged under loa specimens charged without load for 72 h showed a 6% decrease in percentage elon A, which is 29% less compared to the as-delivered condition (Figure 9).The elong was 11%, which was a 50% decrease compared to the as-delivered condition.The analysis of the results for reduction in area (Z) of P91 steel confirmed a d in ductility in the specimens exposed to hydrogen, especially those charged unde The percentage reduction in area (Z) for the specimens charged without load for 72 46.6%, which was a 37% reduction compared to the as-delivered condition.For the mens charged under load for 72 h, the reduction in area was 25.7%, which was decrease compared to the as-delivered condition (Figure 10).An additional static tensile test was conducted on the material under load o for 110 h.The results demonstrate a significant degradation of the material's plast erties.The elongation at fracture (A) was measured at 3.4%, while the reduction (Z) was recorded at 1.6%.These findings indicate a substantial negative impact of gen on the plastic properties of P91 steel, particularly in operating conditions ch ized by sustained load and a hydrogenation environment.A rapid increase in em ment can lead to the swift and uncontrolled failure of the entire structure.
The mechanical tests were complemented by fractographic examinations of t ture surfaces.Such images clearly show surface topography.Examples of fracture s for the material in the as-delivered condition are shown in Figure 11, whereas Fig and 13 show fracture surfaces of specimens tested following hydrogen charging load and under load, respectively.
After the static tensile test, the sample exhibited a ductile fracture with vis crocracks and micropores (Figure 11b).Examination of the sample's side surface r As early as after 24 h, the material's strength decreased to 654 MPa, and after 72 h it decreased to 627 MPa.This confirms that hydrogen affects the mechanical properties of P91 steel, especially during service and operation.
As for the yield strength (R p0.2 ), which is 546 MPa in the as-delivered condition, it decreased as the charging time increased, falling to 501 MPa after 72 h of charging without load (Figure 8).The analysis of the results obtained for the specimens charged under load showed a reduction in yield strength compared to the value obtained for the material in the as-delivered condition (by 20 MPa), but an increase in yield strength compared to the specimens charged without load.The difference in R p0.2 between the specimens charged without load and under load was 13.8 MPa for the charging time of 24 h, and 26.1 MPa for the charging time of 72 h (Figure 8).
The evaluation of plastic properties (A, Z) showed a significant decrease in the ductility of the specimens exposed to hydrogen in comparison with the specimens in the as-delivered condition (Figures 8 and 9).In particular, a considerable deterioration in plastic properties was observed for the specimens that were hydrogen-charged under load.The specimens charged without load for 72 h showed a 6% decrease in percentage elongation A, which is 29% less compared to the as-delivered condition (Figure 9).The elongation A was 11%, which was a 50% decrease compared to the as-delivered condition.
The analysis of the results for reduction in area (Z) of P91 steel confirmed a decrease in ductility in the specimens exposed to hydrogen, especially those charged under load.The percentage reduction in area (Z) for the specimens charged without load for 72 h was 46.6%, which was a 37% reduction compared to the as-delivered condition.For the specimens charged under load for 72 h, the reduction in area was 25.7%, which was a 65% decrease compared to the as-delivered condition (Figure 10).
An additional static tensile test was conducted on the material under load of 30 kN for 110 h.The results demonstrate a significant degradation of the material's plastic properties.The elongation at fracture (A) was measured at 3.4%, while the reduction in area (Z) was recorded at 1.6%.These findings indicate a substantial negative impact of hydrogen on the plastic properties of P91 steel, particularly in operating conditions characterized by sustained load and a hydrogenation environment.A rapid increase in embrittlement can lead to the swift and uncontrolled failure of the entire structure.
The mechanical tests were complemented by fractographic examinations of the fracture surfaces.Such images clearly show surface topography.Examples of fracture surfaces for the material in the as-delivered condition are shown in Figure 11, whereas Figures 12  and 13 show fracture surfaces of specimens tested following hydrogen charging without load and under load, respectively.
After the static tensile test, the sample exhibited a ductile fracture with visible microcracks and micropores (Figure 11b).Examination of the sample's side surface revealed no microcracks or other discontinuities, indicating good mechanical properties and a lack of susceptibility to microcrack formation, which could potentially lead to structural failure.An additional static tensile test was conducted on the material under load of 30 kN for 110 h.The results demonstrate a significant degradation of the material's plastic prop erties.The elongation at fracture (A) was measured at 3.4%, while the reduction in are (Z) was recorded at 1.6%.These findings indicate a substantial negative impact of hydro gen on the plastic properties of P91 steel, particularly in operating conditions character ized by sustained load and a hydrogenation environment.A rapid increase in embrittle ment can lead to the swift and uncontrolled failure of the entire structure.
The mechanical tests were complemented by fractographic examinations of the frac ture surfaces.Such images clearly show surface topography.Examples of fracture surface for the material in the as-delivered condition are shown in Figure 11, whereas Figures 1  and 13 show fracture surfaces of specimens tested following hydrogen charging withou load and under load, respectively.
After the static tensile test, the sample exhibited a ductile fracture with visible mi crocracks and micropores (Figure 11b).Examination of the sample's side surface revealed no microcracks or other discontinuities, indicating good mechanical properties and a lack of susceptibility to microcrack formation, which could potentially lead to structural failure Analysis of the results of the static tensile test of P91 steel samples without load did not reveal any significant changes in the mechanical properties of the material after 24 h of sample exposure in the solution environment.The material strength decreased by 10 MPa (Figure 8), and the yield strength decreased by 21 MPa.Significant decreases in area reduction and elongation were observed after 48 h and 76 h of exposure, while the strength of the material remained relatively unchanged.These results suggest that P91 steel exhibits significant resistance to hydrogen embrittlement up to 24 h, however, prolonged exposure to hydrogen leads to increased embrittlement.The yield strength (R p0.2 ) decreased from 546 MPa to 501 MPa after 72 h of exposure to the hydrogenated solution without load.Samples hydrogenated without load for 72 h exhibited a 6% decrease in elongation (A), representing a 29% reduction compared to the initial state.The percentage reduction in area (Z) for the sample hydrogenated without load for 72 h was 46.6%, indicating a 37% decrease compared to the initial state.
Analysis of fractures and microstructure on flat longitudinal metallographic sections reveals the presence of typical HELP phenomenon transcrystalline microcracks, attributed to a reduction in the strength of the material's micro-regions (Figure 12a,b).Additionally, cracks are visible along grain boundaries.This is a phenomenon associated with hydrogen accumulation in chromium carbide precipitates and the grain boundary area (HEDE) (Figure 12c).The fracture surface also exhibits several defects described as "fish eyes" (Figure 12d), characteristic of the AIDE phenomenon and likely caused by hydrogen accumulation at the interface between metallic inclusions and the matrix.The obtained data confirm that P91 steel exhibits susceptibility to hydrogen embrittlement, consistent with the observed decline in mechanical properties.Analysis of the results of the static tensile test of P91 steel samples without load did not reveal any significant changes in the mechanical properties of the material after 24 h of sample exposure in the solution environment.The material strength decreased by 10 MPa (Figure 8), and the yield strength decreased by 21 MPa.Significant decreases in area reduction and elongation were observed after 48 h and 76 h of exposure, while the strength of the material remained relatively unchanged.These results suggest that P91 steel exhibits significant resistance to hydrogen embrittlement up to 24 h, however, prolonged exposure to hydrogen leads to increased embrittlement.The yield strength (Rp0.2) decreased from 546 MPa to 501 MPa after 72 h of exposure to the hydrogenated solution without load.Samples hydrogenated without load for 72 h exhibited a 6% decrease in elongation (A), representing a 29% reduction compared to the initial state.The percentage reduction in area (Z) for the sample hydrogenated without load for 72 h was 46.6%, indicating a 37% decrease compared to the initial state.
Analysis of fractures and microstructure on flat longitudinal metallographic sections Fractographic analysis of the surfaces indicates a mixed ductile-brittle fracture mode in all cases.A detailed examination of the fracture surfaces of samples subjected to hydrogen charging under load revealed numerous effects of structural degradation due to hydrogen interaction (Figure 13).Numerous "fish-eye" cracks were observed on the fracture surface of the sample after 72 h of exposure to a hydrogen-charging environment under a constant load (Figure 13b).These cracks were so numerous that they overlapped, leading to the formation of microcracks and even larger cracks perpendicular to the tensile direction.It is a typical effect of hydrogen embrittlement, leading to a loss of material cohesion.Example observations of these microcracks are shown in Figure 13c, where the smooth surface of the microcrack, caused by hydrogen interaction, is visible.This phenomenon is described in the literature as HELP and AIDE [25,26,28].Additionally, numerous voids and cracks caused by the presence of hydrogen were observed in the structure.An example of the observed structure from the fracture area with visible hydrogen voids and microcracks along the grain boundaries is shown in Figure 13c.constant load (Figure 13b).These cracks were so numerous that they overlapped, leading to the formation of microcracks and even larger cracks perpendicular to the tensile direction.It is a typical effect of hydrogen embrittlement, leading to a loss of material cohesion.Example observations of these microcracks are shown in Figure 13c, where the smooth surface of the microcrack, caused by hydrogen interaction, is visible.This phenomenon is described in the literature as HELP and AIDE [25,26,28].Additionally, numerous voids and cracks caused by the presence of hydrogen were observed in the structure.An example of the observed structure from the fracture area with visible hydrogen voids and microcracks along the grain boundaries is shown in Figure 13c.
Further metallographic analysis of cross-sections transverse to the tensile direction revealed numerous cracks on the sample surface, which is associated with the interaction of hydrogen according to the mechanism described as HEDE [25,27].Further metallographic analysis of cross-sections transverse to the tensile direction revealed numerous cracks on the sample surface, which is associated with the interaction of hydrogen according to the mechanism described as HEDE [25,27].
In conclusion, it can be positively stated that a hydrogen-containing environment has a very negative effect on the mechanical properties of P91 steel.With long hydrogen exposure times under load, brittleness sharply increases, which is very dangerous from the operational point of view.An important factor affecting the material's behavior in a corrosive, hydrogen containing environment is the additional stress present during hydrogen charging.This is particularly important because these conditions are the closest to the actual operating conditions of most industrial facilities.In this study, a preload of 70% of R p0.2 was applied during the hydrogen charging process, which simulated extremely dangerous conditions.

Figure 2 .
Figure 2. Martensitic structure of P91 steel in the as-delivered condition.

Figure 3 .
Figure 3. Specimen for testing the effect of hydrogen on the mechanical properties of P91 s

Figure 2 .
Figure 2. Martensitic structure of P91 steel in the as-delivered condition.

Figure 2 .
Figure 2. Martensitic structure of P91 steel in the as-delivered condition.

Figure 3 .
Figure 3. Specimen for testing the effect of hydrogen on the mechanical properties of P91 s

Figure 3 .
Figure 3. Specimen for testing the effect of hydrogen on the mechanical properties of P91 steel.

Figure 4 .
Figure 4. Stand for hydrogen charging under load.

Figure 4 .
Figure 4. Stand for hydrogen charging under load.

Figure 5 .
Figure 5.Comparison of the static tensile strength of P91 steel in the as-delivered condit following hydrogen charging without preloading.

Figure 5 .
Figure 5.Comparison of the static tensile strength of P91 steel in the as-delivered condition and following hydrogen charging without preloading.

Figure 6 .
Figure 6.Comparison of the static tensile strength of P91 steel in the as-delivered following hydrogen charging under preload.

Figure 6 .
Figure 6.Comparison of the static tensile strength of P91 steel in the as-delivered condition and following hydrogen charging under preload.

Figure 7 .
Figure 7. Tensile strength comparison for the different variants of the hydrogen charging te

Figure 7 .
Figure 7. Tensile strength comparison for the different variants of the hydrogen charging test.

Figure 8 .
Figure 8. Yield strength comparison for the specimens in the as-delivered condition follow drogen charging without load, and following hydrogen charging under a load of 30 kN.

Figure 8 .
Figure 8. Yield strength comparison for the specimens in the as-delivered condition following hydrogen charging without load, and following hydrogen charging under a load of 30 kN.

Figure 9 .
Figure 9. Percentage elongation comparison for the specimens in the as-delivered condition, ing hydrogen charging without load, and following hydrogen charging under a load of 30

Figure 9 .Figure 10 .
Figure 9. Percentage elongation comparison for the specimens in the as-delivered condition, following hydrogen charging without load, and following hydrogen charging under a load of 30 kN. Materials 2024, 17, x FOR PEER REVIEW

Figure 10 .
Figure 10.Percentage reduction in area comparison for the specimens in the as-delivered condition following hydrogen charging without load, and following hydrogen charging under a load of 30 kN.

Figure 10 .
Figure 10.Percentage reduction in area comparison for the specimens in the as-delivered condition following hydrogen charging without load, and following hydrogen charging under a load of 30 kN

Figure 11 .Figure 11 .
Figure 11.Structure of P91 steel in the as-delivered condition: (a) martensitic structure on a fla cross-section; (b) mixed brittle-ductile fracture surface following the static tensile strength test.

Materials 2024 ,
17, x FOR PEER REVIEW 10 of 14 Charging time: 72 h without load

Figure 12 .
Figure 12.Fracture surface after static tensile testing following hydrogen charging without load for 72 h: (a) overall view of the fracture surface; (b) mixed ductile-brittle fracture structure with visible cleavage fracture; (c) intergranular cracks in the fracture zone; (d) fracture surface with a visible "fish-eye" hydrogen crack.

Figure 12 .
Figure 12.Fracture surface after static tensile testing following hydrogen charging without load for 72 h: (a) overall view of the fracture surface; (b) mixed ductile-brittle fracture structure with visible cleavage fracture; (c) intergranular cracks in the fracture zone; (d) fracture surface with a visible "fish-eye" hydrogen crack.

Figure 13 .Figure 13 .
Figure 13.Fracture surface after static tensile testing following hydrogen charging with load for 72 h: (a) overall view of the fracture surface; (b) fracture surface with visible "fish-eye" overlap cracks; (c) microstructure in the fracture zone (flat cross-section), revealing voids and blisters associated with hydrogen presence, with microcracks along grain boundaries; (d) smooth surface of the hydrogen crack on the fracture surface.In conclusion, it can be positively stated that a hydrogen-containing environment has a very negative effect on the mechanical properties of P91 steel.With long hydrogen

Table 1 .
Mechanical properties of P91 steel specimens subjected to static tensile strength tests in the as-delivered condition and following hydrogen charging without preloading.

Table 2 .
Mechanical properties of P91 steel specimens subjected to static tensile strength tes as-delivered condition and following hydrogen charging under load.

Table 2 .
Mechanical properties of P91 steel specimens subjected to static tensile strength tests in the as-delivered condition and following hydrogen charging under load.

ID Charging Time Tensile Strength Rm, MPa Yield Strength R p0.2 , MPa Percentage Elongation A, % Percentage Reduction in Area Z, %
aterials 2024, 17, x FOR PEER REVIEW