Study on Hydrogen Embrittlement Behavior of X65 Pipeline Steel in Gaseous Hydrogen Environment

Abstract

Pipeline steel is highly susceptible to hydrogen embrittlement (HE) in hydrogen environments, which compromises its structural integrity and operational safety. Existing studies have primarily focused on the degradation trends of mechanical properties in hydrogen environments, but there remains a lack of quantitative failure prediction models. To investigate the failure behavior of X65 pipeline steel under hydrogen environments, this paper utilized notched round bar specimens with three different radii and smooth round bar specimens to examine the effects of pre-charging time, the coupled influence of stress triaxiality and hydrogen concentration, and the coupled influence of strain rate and hydrogen concentration on the HE sensitivity of X65 pipeline steel. Fracture surface morphologies were characterized using scanning electron microscopy (SEM), revealing that hydrogen-enhanced localized plasticity (HELP) dominates failure mechanisms at low hydrogen concentrations, while hydrogen-enhanced decohesion (HEDE) becomes dominant at high hydrogen concentrations. The results demonstrate that increasing stress triaxiality or decreasing strain rate significantly intensifies the HE sensitivity of X65 pipeline steel. Based on the experimental findings, failure prediction models for X65 pipeline steel were developed under the coupled effects of hydrogen concentration and stress triaxiality as well as hydrogen concentration and strain rate, providing theoretical support and mathematical models for the engineering application of X65 pipeline steel in hydrogen environments.

1. Introduction

Against the backdrop of profound transformations in the global energy landscape, hydrogen energy has emerged as a pivotal clean energy source and a focal point for future energy development [1]. Pipeline-based hydrogen transportation [2] offers continuous reliability, cost effectiveness, and high energy density, making it ideal for large-scale, long-distance hydrogen delivery. Hydrogen-blended natural gas (HBNG) transportation [3,4,5], which utilizes existing natural gas pipeline networks, not only reduces infrastructure development costs but also bridges hydrogen energy with conventional energy systems, providing a cost-effective transitional strategy for hydrogen adoption. However, the distinct physicochemical properties of hydrogen compared to natural gas—particularly its smaller molecular size and higher diffusivity—enable hydrogen atoms to permeate metallic microstructures more readily, accelerating premature pipeline material degradation [6]. This underscores the critical need to investigate the hydrogen compatibility of in-service pipeline materials, ensuring the safety and feasibility of HBNG transportation systems.

Current research on hydrogen compatibility of pipeline steels predominantly employs electrochemical hydrogen charging to simulate material behavior in hydrogen environments [7,8,9,10,11]. However, this laboratory-accelerated charging method fundamentally diverges from real-world service conditions where hydrogen ingress primarily originates from corrosion reactions or high-pressure transportation environments. Critical parameters such as hydrogen concentration gradients, permeation kinetics, and stress–hydrogen synergistic effects under actual operational scenarios exhibit intrinsic differences from those in electrochemical charging. These fundamental discrepancies may lead to deviations between laboratory-derived conclusions and practical failure mechanisms, potentially compromising the reliability of hydrogen compatibility assessments for engineering applications.

Existing studies predominantly focus on the hydrogen embrittlement (HE) susceptibility of high-strength pipeline steels such as X70, X80, and X100 [12,13,14]. Asadipoor et al. [12] demonstrated that hydrogen exposure induces fracture mode transitions in X70 steel, shifting from ductile dimple rupture to cleavage-dominated failure. Similar hydrogen-induced fracture mechanism alterations have been observed in X100 and X80 steels by Nanninga et al. [15] and Zhou et al. [16], respectively. The ductile-to-brittle fracture mode transition under hydrogen embrittlement conditions was also reported in X65 pipeline steel by Lin et al. [17] and Mendibide et al. [18]. Regarding hydrogen’s impact on mechanical properties, conflicting findings persist: some studies report reductions in yield strength and tensile strength [19,20,21], while others suggest no significant impact on these strength parameters [16]. This divergence highlights the absence of consensus on hydrogen’s fundamental role in mechanical degradation, underscoring the need for further mechanistic investigations.

Upon ingress into metallic structures, hydrogen primarily influences ductility and may induce premature failure, particularly in aging pipelines with pre-existing material degradation [22,23]. To evaluate hydrogen’s detrimental effects on ductility, slow strain rate tensile (SSRT) testing of smooth or notched round bar specimens is typically employed [13,24,25,26]. However, conducting such hydrogen-environment experiments involves substantial technical challenges and prohibitive costs, leading most studies to rely predominantly on finite element (FE) simulations for mechanistic analysis [27].

Extensive research has been conducted on the failure behavior of notched round bar specimens under tensile loading in air environments [28]. Rice [29] established the relationship between critical plastic strain and stress triaxiality by analyzing the expansion of a single spherical void in ideal elastic-plastic materials. Kanvinde et al. [30] proposed a stress-modified critical strain model to predict fracture initiation under multiaxial stress states. Yao [31] identified the first principal stress and stress triaxiality as dominant factors governing tensile fracture and subsequently developed stress-based critical fracture criteria and corresponding predictive models. Huang et al. [32] integrated numerical simulations with optimization algorithms to analyze tensile test data from various notched specimens, deriving a generalized failure function for notch-induced fracture.

In summary, research on the effects of stress triaxiality and strain rate on the hydrogen embrittlement behavior of X65 pipeline steel remains limited, and quantitative failure models are still lacking. To address this gap, this study conducted tensile tests in hydrogen environments and established failure models for X65 pipeline steel under hydrogen concentration. These models provide a scientific basis for the safety assessment and structural design of X65 pipeline steel used in hydrogen transportation.

2. Materials and Methods

2.1. Materials

All specimens used in this study were extracted from newly produced X65 pipeline steel, whose yield strength and tensile strength are 486 MPa and 597 MPa, respectively. The material exhibits a typical microstructure (Figure 1) consisting of ferrite and pearlite phases, indicative of its favorable strength–toughness balance.

Table 1. Average chemical composition of the investigated steel in weight percentage.

Materials	C	Si	Mn	P	S	Cr	Nb	Ti	Al
API 5L X65	0.076	0.22	1.52	0.015	0.0034	0.18	0.044	0.011	0.022

lists the primary chemical composition of the steel, aligning with API 5L specifications for X65-grade pipeline applications.

2.2. Experimental Methods

The specimen geometries employed in this study are illustrated in Figure 2, including (a) smooth round bar, (b) notched round bar with R1.2 radius, (c) notched round bar with R2.3 radius, and (d) notched round bar with R8 radius. Based on Bridgman’s analytical formula for stress triaxiality calculation [33], the initial stress triaxiality values of these four specimens were determined as 0.33, 1.14, 0.84, and 0.51, respectively.

To evaluate hydrogen-induced mechanical degradation, the experimental protocol comprised three sequential stages:

• Hydrogen Pre-Charging

Specimens were exposed to hydrogen-nitrogen gas mixtures (25 vol% H₂ + 75 vol% N₂, 50 vol% H₂ + 50 vol% N₂, and 75 vol% H₂ + 25 vol% N₂) at 10 MPa pressure to achieve hydrogen ingress. In the study of hydrogen pre-charging effects, the hydrogen pre-charging durations for the specimens were set to 0 h (no charge), 1 h, 4 h, 7 h, and 24 h. When investigating the effects of stress triaxiality and strain rate, all specimens were pre-charged with hydrogen for 24 h. The hydrogen pre-charging test of all specimens was carried out in a static kettle on the same equipment (HSC-100E, Bairoe, Shanghai, China).

2.

Hydrogen Quantification

The hydrogen desorption rate as a function of temperature was characterized via thermal desorption spectroscopy (TDS, HTDS-003, R-DEC Co., Ltd., Tsukuba, Japan). Based on the distinct temperature ranges associated with reversible and irreversible hydrogen trapping, the reversible hydrogen content in the specimens was quantified.

3.

In Situ Tensile Testing

Pre-charged specimens underwent slow strain rate tensile (SSRT) testing in a hydrogen-blended environment (10 MPa chamber pressure, 20 °C). The experimental conditions of each in situ tensile specimen were exactly the same as those during the pre-charging of hydrogen. When studying the effect of strain rate, the strain rates of the samples were 1 × 10⁻⁵/s, 1 × 10⁻⁴/s, and 1 × 10⁻³/s, respectively; the strain rate of all samples was 1 × 10⁻⁴/s when studying the effect of stress triaxiality. The in situ tensile test of all specimens was performed in a dynamic kettle on the same equipment (HSC-100E).

3. Results and Discussion

3.1. Hydrogen Pre-Charge

Figure 3 presents the results of in situ slow strain rate tensile (SSRT) tests on smooth round bar specimens under varying hydrogen pre-charging durations. The specimen tested in a hydrogen-free environment (black curve) exhibited the maximum elongation, confirming that X65 pipeline steel retains its optimal plastic deformation capability in the absence of hydrogen. In contrast, even non-pre-charged specimens exposed to hydrogen environments showed a reduction in elongation by approximately 1 mm compared to hydrogen-free conditions, corresponding to an 11% hydrogen-induced elongation loss. This marked degradation in deformation capacity underscores the severe detrimental impact of hydrogen embrittlement on the material’s plasticity. The consistency of elongation reduction across all hydrogen-exposed specimens, regardless of pre-charging time, further validates hydrogen’s dominant role in suppressing dislocation mobility and promoting brittle fracture mechanisms.

With increasing pre-charging duration, both uniform elongation and fracture displacement of X65 pipeline steel in the plastic deformation stage progressively decreased. The ultimate load was reached earlier compared to hydrogen-free conditions; however, no definitive trend in ultimate load magnitude was observed with prolonged pre-charging time. When the pre-charging duration reached 24 h (yellow solid line in Figure 3), the load–displacement curve underwent a pronounced alteration: a rapid load drop occurred shortly after peak load attainment, with fracture displacement decreasing from 9.15 mm (hydrogen-free) to 5.83 mm, while fracture load increased from 9322 N to 13,971 N (as shown in

Table 2. Fracture displacement and fracture load at different pre-charge duration.

Pre-Charge Duration	Fracture Displacement (mm)	Fracture Load (N)
N2	9.15	9322
0 h (No charge)	8.23	9375
1 h	8.07	10,286
4 h	7.30	11,007
7 h	6.86	10,321
24 h	5.83	13,971

). This significant reduction in fracture displacement indicates severe ductility deterioration, as specimens fractured prematurely after necking without sufficient plastic deformation.

Following the tensile tests, the fracture surface dimensions and post-fracture elongation were measured to determine reduction of area (RA) and elongation after fracture (EL), as summarized in Figure 4a. To evaluate HE susceptibility, two standardized metrics were employed:

$$REL={\displaystyle \frac{{EL}_{N2}-{EL}_{H2}}{{EL}_{N2}}}\times 100$$

(1)

where REL is the relative elongation, ${EL}_{N2}$ is the elongation of the sample after break in the nitrogen environment, and ${EL}_{H2}$ is the elongation after break of the sample in the hydrogen environment.

$$RRA={\displaystyle \frac{{RA}_{N2}-{RA}_{H2}}{{RA}_{N2}}}\times 100$$

(2)

where RRA is relative reduction of area, and ${RA}_{N2}$ and ${RA}_{H2}$ represent the reduction in area under the nitrogen environment and hydrogen environment.

Figure 4a demonstrates that increasing pre-charging duration leads to marked reductions in both RA and EL, aligning with the established HE degradation pattern for X65 pipeline steel. Figure 4b reveals the evolution of hydrogen susceptibility indices (RRA and REL) with pre-charging time: under shorter durations (0 h, 1 h, 4 h, and 7 h), both RRA and REL exhibit gradual increases—rising by 4% and 10%, respectively, from 0 h to 7 h. Notably, a sharp escalation occurred at 24 h pre-charging, where REL and RRA surged by 20% and 27%, reflecting intensified hydrogen damage accumulation.

Figure 5a–c displays the fracture morphology of X65 pipeline steel specimens tested at a strain rate of 1 × 10⁻⁵/s in N₂, characterized by scanning electron microscopy (SEM). The equipment model of the SEM is VEGAIIXMH. According to ASTM E1820 [34] and ISO 12135 [35], the macroscopic fracture surface exhibits a classic cup-and-cone morphology (Figure 5a) indicative of ductile failure, with distinct shear lips inclined at approximately 45° to the loading axis at the specimen periphery, reflecting localized plastic shear deformation during post-necking stages. High-magnification imaging of the central fracture zone (Figure 5b, magnified from the red frame in Figure 5a) revealed uniformly distributed equiaxed microvoids, demonstrating isotropic void nucleation and growth during plastic deformation, consistent with ductile crack propagation from the specimen center outward. The lateral surface morphology (Figure 5c) further corroborates extensive plastic deformation, displaying smooth profiles with pronounced necking contours. Collectively, the cup-and-cone geometry, shear lips, homogeneous microvoid distribution, and necking evidence conclusively validate that fracture in nitrogen environments occurs primarily through ductile mechanisms.

Figure 5d–f depicts the hydrogen-induced fracture morphological evolution in X65 steel specimens subjected to 24 h hydrogen pre-charging in a 50 vol% H₂ environment, followed by in situ tensile testing at a strain rate of 1 × 10⁻⁵/s. The reduced fibrous zone area (Figure 5d) indicates that hydrogen suppressed the ductile fracture process in the central region via the HELP mechanism, which accelerated dislocation motion and promoted localized shear band formation. Concurrently, hydrogen adsorption reduced surface energy, facilitating outward extension of shear bands through HELP-driven plasticity, thereby enlarging the shear lip area. Additionally, the fracture surface exhibited a stepped distribution of tear ridges, a feature closely linked to hydrogen’s dynamic adsorption–desorption behavior at crack tips.

Figure 5e (which shows the magnified area marked by red frames in Figure 5d) illustrates the formation of secondary dimples, a hallmark of hydrogen-induced microvoid coalescence, facilitated by hydrogen’s localized plasticity synergy. When hydrogen concentration is low, the HELP mechanism promotes dislocation slip around microvoids, enabling void growth and linkage under reduced stress. Furthermore, the circumferential cracks and irregular fracture surfaces (Figure 5f) reveal hydrogen’s multiscale damage modes: at elevated hydrogen concentrations, the HEDE mechanism dominates, causing hydrogen segregation at grain boundaries, which weakens grain boundary cohesion energy and promotes intergranular fracture. Concurrently, the coupling between stress gradients and hydrogen diffusion elevates surface hydrogen concentrations, initiating surface crack nucleation [36]. As hydrogen diffuses inward and synergizes with stress fields, cracks propagate radially from the surface, ultimately forming a layered fracture structure characterized by a brittle surface layer and ductile core (Figure 5f) [19]. This dual mechanism—HELP-driven plasticity localization and HEDE-governed interfacial embrittlement—collectively dictates the hierarchical failure morphology under hydrogen exposure.

In both AISI 1020 [37] and AISI 4135 [36] steels, it was observed that hydrogen induces premature failure of specimens, accompanied by a transition in fracture morphology from dimpled ductile fracture to cleavage-like brittle fracture.

3.2. Notch Tensile with Hydrogen

Figure 6 displays the load–displacement curves of notched round bar specimens (R8, R2.3, and R1.2) after 24 h hydrogen pre-charging and subsequent in situ tensile testing at a strain rate of 1 × 10⁻⁴/s under hydrogen-blended environments with 25 vol% H₂, 50 vol% H₂, and 75 vol% H₂. In the initial elastic stage (displacement less than 1.0 mm), all curves exhibited nearly identical slopes, indicating negligible hydrogen-induced alterations to the Young’s modulus during initial deformations. As displacement increased, all specimens reached their ultimate load at approximately 1 mm displacement. Beyond this point, distinct post-necking behaviors emerged depending on stress triaxiality: the R8 specimen (lowest triaxiality) demonstrated the highest ductility with extensive plastic deformation, followed by the R2.3 specimen (moderate triaxiality) showing intermediate ductility, while the R1.2 specimen (highest triaxiality) exhibited the poorest ductility, characterized by a sharp load drop immediately after peak load. This triaxiality-dependent ductility hierarchy persisted across all hydrogen concentrations, highlighting the dominant role of stress state in governing hydrogen-affected fracture behavior.

Hydrogen-induced ductility degradation manifested systematically across all tested specimens. In the 25% hydrogen environment (Figure 6a), the R8 notched specimen achieved an elongation of 2.90 mm, significantly higher than the 2.15 mm for R2.3 and 1.75 mm for R1.2 specimens. Elevating hydrogen concentration to 50% (Figure 6b) reduced elongations to 2.60 mm, 2.05 mm, and 1.57 mm, with retained yet diminished ductility. Under 75% hydrogen exposure (Figure 6c), elongations further decreased to 2.50 mm, 2.00 mm, and 1.50 mm for R8, R2.3, and R1.2, respectively. The progressive 13.8–14.3% elongation reduction per 25% hydrogen increment quantitatively demonstrates hydrogen’s severe embrittling effect on X65 steel, particularly pronounced in high-stress-triaxiality geometries where hydrogen-assisted crack initiation dominates plastic deformation.

The HE index can be expressed in terms of relative reduction of area:

$$HE={\displaystyle \frac{{RA}_{N_2}-{RA}_{H_2}}{{RA}_{N_2}}}\times 100$$

(3)

where ${RA}_{N_2}$ is the reduction in area of the sample in the nitrogen environment, and ${RA}_{H_2}$ is the reduction in area of the sample in the hydrogen environment.

The ductility indicators and HE index of the slow-strain tensile test in a hydrogen environment are shown in Figure 7. Figure 7a reveals a pronounced increase in ultimate load with rising stress triaxiality. When stress triaxiality escalated from 0.51 to 1.14, the ultimate load increased by approximately 7000 N. This enhancement is attributed to the constrained plastic zone under elevated stress triaxiality, which elevates the material’s load-bearing capacity. However, under identical stress triaxiality conditions, hydrogen concentration exhibited no consistent influence on ultimate load, suggesting that stress state—rather than hydrogen embrittlement—dominates ultimate load regulation. This conclusion aligns with findings from multiple studies [17,38], reinforcing the primary role of mechanical constraints over hydrogen-induced degradation in governing load limits.

Figure 7b demonstrates a declining trend in the reduction in area (RA) with increasing stress triaxiality, indicating suppressed plastic deformation capacity under high-triaxiality conditions. Hydrogen concentration further exacerbates this reduction: RA decreased by approximately 10% when hydrogen content increased from 25% to 50%, followed by an additional 10% reduction as hydrogen concentration rose to 75%. This cumulative 20% RA loss highlights the severity of HE in high-concentration environments, where HEDE and HELP synergistically override the material’s intrinsic ductility.

Figure 7c quantifies the HE index of specimens under hydrogen environments. The HE index increased significantly with hydrogen concentration. For the R8 specimen, the HE index rose from 14% at 25% hydrogen to approximately 28% at 50% hydrogen—a 14% increase—and further surged to 40% at 75% hydrogen. Under identical hydrogen exposure, higher stress triaxiality amplified HE severity. At 75% hydrogen concentration, the R1.2 specimen exhibited a 15% higher HE index than R8, demonstrating that elevated stress triaxiality exacerbates hydrogen-induced degradation. The same trend was observed in X70 pipeline steel [38] and AISI 4135 steel [36] in the ex situ hydrogen filling test. This synergistic interaction between hydrogen concentration and stress triaxiality aligns with micromechanical models of hydrogen-enhanced damage accumulation under constrained plasticity.

In summary, ultimate load is predominantly governed by stress triaxiality, while material ductility is primarily influenced by hydrogen concentration. Increasing hydrogen concentration intensifies plastic damage, as evidenced by marked reductions in both reduction of area (RA) and elongation, confirming HE as the dominant mechanism driving plastic degradation. Furthermore, the synergy between HE and stress triaxiality exacerbates material embrittlement—elevated stress triaxiality accelerates hydrogen-assisted damage accumulation, promoting premature brittle fracture. This interaction highlights the critical role of multiaxial stress states in amplifying hydrogen’s detrimental effects.

Under the same hydrogen environment (75 vol% H₂), the fracture morphologies of specimens with different stress triaxialities exhibit significant differences (Figure 8). For specimens with low stress triaxiality (R2.3, Figure 8a,b), the fracture surface shows higher roughness and a relatively smaller area, with prominent shear lip features visible along the fracture edges. Microscopic analysis in Figure 8b (which shows the magnified area marked by red frames in Figure 8a) reveals that the fracture surface is primarily composed of numerous small, shallow dimples, accompanied by a few larger, deeper dimples; smooth cleavage facets are observed at the bottom of certain regions. This indicates that the fracture of X65 pipeline steel at this stage is predominantly governed by a ductile mechanism involving microvoid nucleation and coalescence, though a local tendency toward a ductile-to-brittle transition is evident.

In contrast, specimens with high stress triaxiality (Figure 8c,d) display a typical brittle fracture mode. The macroscopic fracture surface appears flat, with clear tear ridges observed in the central region (Figure 8c). As shown in Figure 8d (which shows the magnified area marked by red frames in Figure 8c), the dimple structure has almost completely disappeared, replaced by large, continuous cleavage facets, accompanied by intergranular secondary cracks (indicated by red arrows in Figure 8d). This suggests that the hydrostatic stress gradient induced by high stress triaxiality significantly promotes the diffusion and accumulation of hydrogen atoms toward the crack tip, leading to the dominance of the HEDE mechanism and ultimately resulting in macroscopic brittle fracture behavior of the material.

In order to accurately and quantitatively predict the failure behavior of X65 pipeline steel under different hydrogen concentrations and stress triaxialities, it is necessary to establish a failure prediction model. In the current literature, the relationship between fracture strain and stress triaxiality has been extensively explored through theoretical models and experimental studies [39,40,41,42]. Building upon these foundations, the failure criterion selected in this work is as followed:

$${\epsilon }_f={\mathrm{A}}_1+{\mathrm{A}}_2\mathrm{e}\mathrm{x}\mathrm{p}(-\eta /{\mathrm{A}}_3)$$

(4)

where ${\epsilon }_f$ is fracture strain, η is the stress triaxiality, and ${\mathrm{A}}_1, {\mathrm{A}}_2$, and ${\mathrm{A}}_3$ are constants.

Figure 9 shows the best-fit curves derived from the uncharged test data for each specimen geometry.

$${\epsilon }_f=1.05+3.38\mathrm{e}\mathrm{x}\mathrm{p}(-\eta /0.16)$$

(5)

To incorporate the hydrogen concentration effect into the failure model, the fracture strains were normalized (see Figure 10). This approach aligns with the findings of Sofronis et al. [43], who identified reversible hydrogen as the primary contributor to hydrogen embrittlement. Consequently, all hydrogen concentrations referenced in this study specifically pertain to reversible hydrogen concentrations, which dominate the degradation mechanism under the tested conditions.

Figure 10 shows that the fracture strain of X65 pipeline steel after normalization has an obvious linear relationship with the reversible hydrogen concentration. Therefore, combined with Equation (5), the failure model of X65 pipeline steel with respect to stress triaxiality and hydrogen concentration can be obtained:

$${\epsilon }_f=\left[1.05+3.38\mathrm{e}\mathrm{x}\mathrm{p}(-\eta /0.16)\right](1-1.55C_H)$$

(6)

where $C_H$ represents the reversible hydrogen concentration.

Equation (6) was verified using the experimental results, as shown in Figure 11.

Figure 11 compares experimental results with predictions from the failure model under hydrogen concentrations of 0.054, 0.074, and 0.210 wppm (weight parts per million, quantified via TDS). The mean squared error (MSE) values between experimental and simulated fracture strains are 0.0046, 0.0029, and 0.0009, respectively. These minimal deviations demonstrate the strong predictive capability of the failure model in Equation (6) for estimating fracture strain across varying hydrogen concentrations and stress triaxiality levels. This alignment between predicted and experiment shows the model’s utility for safety assessments of X65 pipelines in hydrogen-blended environments.

3.3. Strain Rate

Figure 12 presents the load–displacement curves from tensile tests conducted under coupled effects of strain rate and hydrogen concentration. Experiments were performed in nitrogen and hydrogen-blended environments (after 24 h pre-charge in 25 vol%, 50 vol%, and 75 vol% H₂) at strain rates of 1 × 10⁻³/s, 1 × 10⁻⁴/s, and 1 × 10⁻⁵/s to systematically evaluate hydrogen–strain rate synergism.

Figure 12a illustrates the tensile test results under nitrogen environments, demonstrating the significant influence of strain rate on mechanical performance. At a strain rate of 1 × 10⁻³/s, the fracture displacement was 7.5 mm, increasing to 8 mm at 1 × 10⁻⁴/s and further to 9 mm at 1 × 10⁻⁵/s. The gradual rise in fracture displacement with decreasing strain rate arises from enhanced dislocation slip and rearrangement under prolonged deformation times, which alleviates stress concentration and delays the failure process. The observed trend aligns with classical dislocation dynamics theory, where reduced strain rates enable more uniform plastic flow and microvoid coalescence, thereby improving ductility [44].

In the low-hydrogen environment (25 vol% H₂, Figure 12b), the X65 retained significant ductility with elongations of 8 mm, 7.5 mm, and 7.2 mm at strain rates of 1 × 10⁻³/s, 1 × 10⁻⁴/s, and 1 × 10⁻⁵/s, respectively, as hydrogen’s detrimental influence remained partially mitigated by limited hydrogen–defect interactions. In the moderate hydrogen concentration (50 vol% H₂, Figure 12c), ductility declined markedly due to intensified HE, where HEDE at grain boundaries and inclusions accelerated damage initiation. Under high hydrogen exposure (75 vol% H₂, Figure 12d), hydrogen permeation depth increased significantly, promoting extensive interactions with microstructural defects (e.g., grain boundaries and inclusions) and triggering rapid damage accumulation. At the lowest strain rate (1 × 10⁻⁵/s), elongation plummeted to 5.5 mm, reflecting the dominance of hydrogen–stress synergy in suppressing dislocation-mediated plasticity and amplifying brittle fracture mechanisms.

Strain rate governs material failure behavior by modulating the dynamic competition between hydrogen diffusion and plastic deformation. Under high-strain-rate conditions (1 × 10⁻³/s), rapid loading restricts hydrogen diffusion within X65 steel, confining hydrogen accumulation to near-surface regions and shallow traps with low binding energy. In this regime, HE is partially suppressed, and the HELP mechanism dominates—promoting localized dislocation motion while minimally affecting macroscopic strength, allowing plasticity-driven deformation to prevail. In contrast, at low strain rates (1 × 10⁻⁵/s), prolonged deformation enables hydrogen to diffuse into deep traps such as grain boundaries and precipitates, significantly elevating internal hydrogen concentrations. This facilitates the transition to HEDE dominance, where hydrogen weakens interatomic bonding forces, drastically reducing both strength and ductility. The synchronization of crack propagation with hydrogen diffusion kinetics under low strain rates promotes cleavage-type brittle fracture, characterized by abrupt load drops and minimal plastic deformation. This strain rate-dependent interplay between HELP and HEDE mechanisms underscores hydrogen’s dual role in either exacerbating localized plasticity or triggering interfacial embrittlement, contingent on the relative timescales of hydrogen migration and mechanical loading.

Figure 13 summarizes the calculated RA and HE index derived from post-test measurements of fracture surface dimensions across all specimens.

In N₂ (Figure 13a), increased strain rates (e.g., from 1 × 10⁻⁵/s to 1 × 10⁻³/s) suppress microcrack propagation by limiting deformation time, resulting in about a 10% RA decrease, as insufficient plasticity prevents stress relaxation. Conversely, in hydrogen environments, faster strain rates reduce hydrogen embrittlement severity: RA increases by 5–20% with strain rate elevation (from 1 × 10⁻⁵/s to 1 × 10⁻³/s), as shortened test durations (less than 10 min at 1 × 10⁻³/s) restrict hydrogen ingress, minimizing hydrogen–defect interactions. This inverse trend is corroborated by the HE index in Figure 13b, where the HE index decreases progressively with strain rate acceleration. At 1 × 10⁻³/s, the HE index converges across hydrogen concentrations, differing by 10%, due to insufficient hydrogen permeation depth to activate bulk embrittlement mechanisms. These findings underscore strain rate’s dual role: in inert environments, it governs plasticity kinetics; in hydrogen, it modulates hydrogen availability for embrittlement via diffusion–time competition, offering practical guidelines for optimizing deformation rates to mitigate hydrogen damage in pipeline operations.

In summary, elevated hydrogen concentrations significantly exacerbate the brittle fracture propensity of X65 pipeline steel during tensile deformation, particularly under low-strain-rate conditions. This degradation is predominantly governed by sufficient hydrogen diffusion and subsequent interactions with microstructural features such as grain boundaries and inclusions, which amplify HE through HEDE mechanisms [45]. Conversely, high strain rates partially suppress HE by restricting hydrogen diffusion kinetics, thereby limiting hydrogen accumulation at critical defect sites. Under rapid loading (e.g., 1 × 10⁻³/s), the dominance of HELP mechanisms promotes dislocation-mediated plasticity, preserving residual ductility even in hydrogen-rich environments. These dual dependencies on hydrogen concentration and strain rate underscore the necessity of incorporating hydrogen–strain rate synergism into integrity assessments for pipelines operating in hydrogen-blended gas systems.

To better guide the engineering application of pipeline steels in hydrogen environments, this study established a failure prediction model for X65 steel under the coupled effects of strain rate and hydrogen concentration. The research first systematically analyzed fracture behavior under single-variable conditions to provide foundational data and theoretical support for subsequent investigations of coupling mechanisms.

The initial phase focused on characterizing the relationship between fracture strain and strain rate for X65 steel in hydrogen-free conditions. A series of tensile fracture experiments were conducted across strain rates ranging from 1 × 10⁻⁵/s to 1 × 10⁻³/s. The resulting fracture strain data are summarized in Figure 14.

As can be seen from Figure 14, in the absence of hydrogen, the relationship between the fracture strain and strain rate is as follows:

$${\epsilon }_{f,noH}=2.26+0.33\mathrm{l}\mathrm{g}\left(\dot{\epsilon }\right)+0.03{\mathrm{l}\mathrm{g}}^2\left(\dot{\epsilon }\right)$$

(7)

where ${\mathsf{\epsilon }}_{\mathrm{f},\mathrm{n}\mathrm{o}\mathrm{H}}$ is the fracture strain without hydrogen, and $\dot{\epsilon }$ is the strain rate.

When considering the degradation of the mechanical properties of X65 pipeline steel in the hydrogen environment, the following two types of degradation functions are compared.

Type 1:

$${\epsilon }_f={\epsilon }_{f,noH}(\mathrm{A}+\mathrm{B}{C}_H)$$

(8)

Type 2:

$${\epsilon }_f={\epsilon }_{f,noH}\left(\mathrm{E}{C_H}^{\mathrm{F}}\right)$$

(9)

where ${\mathsf{\epsilon }}_{\mathrm{f}}$ is fracture strain with hydrogen, and A, B, E, and F are constants.

After trial and adjustment, A, B, E, and F were finally determined to be 0.72, −1.12, 0.37, and −0.22, respectively, and the comparison between model 1 and model 2 and the test results is shown in Figure 15.

Table 3. Two types of failure model for X65 pipeline steel under hydrogen.

Type	Function	MSE
Type 1	${\epsilon }_f={\epsilon }_{f,noH}(\mathrm{A}+\mathrm{B}{C}_H)$	0.0228
Type 2	${\epsilon }_f={\epsilon }_{f,noH}(\mathrm{E}{C_H}^{\mathrm{F}})$	0.0088

shows that the mean square error (MSE) of type 2 is 0.0088, which is much lower than the error of type 1 of 0.0228, so it can be determined that the fracture failure model of X65 pipeline steel under the coupling of strain rate and hydrogen concentration is ${\epsilon }_f=(2.26+0.33\mathrm{l}\mathrm{g}(\dot{\epsilon })+0.03{\mathrm{l}\mathrm{g}}^2(\dot{\epsilon }))(0.37{C_H}^{-0.22})$.

4. Conclusions

To investigate the effects of stress triaxiality on the fracture behavior of X65 pipeline steel under hydrogen environments, this study conducted pre-charging and in situ tests on both notched and smooth round bar specimens under varying hydrogen concentrations. The key conclusions are as follows:

(1)

The presence of hydrogen weakens the plastic deformation capacity of the material, resulting in a significant decrease in elongation and a failure mode from ductile fracture to brittle fracture. The increase of hydrogen pre-charge duration will weaken the plastic deformation ability of X65 pipeline steel;

(2)

The synergistic effect of hydrogen and stress triaxiality affect the failure process of the material, and the high stress triaxiality will aggravate the influence of HE effect on the fracture behavior of X65 pipeline steel;

(3)

Under the coupling effect of strain rate and hydrogen concentration, the slower the strain rate and the higher the hydrogen concentration, the higher the hydrogen enrichment under the action of the stress field, which leads to the more serious HE of X65 pipeline steel;

(4)

Through the analysis of the experimental results in the hydrogen environment, the failure model under the coupling effect of stress triaxiality and hydrogen concentration is ${\epsilon }_f=\left[1.13+0.614\mathrm{e}\mathrm{x}\mathrm{p}(-\eta /0.34)\right](1-1.75C_H)$. The failure model under the coupling of strain rate and hydrogen concentration is ${\epsilon }_f=(2.26+0.33\mathrm{l}\mathrm{g}(\dot{\epsilon })+0.03{\mathrm{l}\mathrm{g}}^2(\dot{\epsilon }))(0.37{C_H}^{-0.22})$.