CTOD Evaluation of High-Nitrogen Steels for Low-Temperature Welded Structures

Abstract

Welded structures, such as offshore platforms, require robust toughness in their heat-affected zones (HAZ) to withstand low-temperature environments. The coarse-grained HAZ (CGHAZ) adjacent to the fusion boundary often exhibits reduced toughness due to grain coarsening, particularly under high heat input welding conditions aimed at enhancing productivity. To address this, high-nitrogen steels containing TiN particles were developed to suppress austenite grain growth by leveraging the thermal stability of TiN precipitates. Three high-nitrogen steels with varying carbon contents (0.09%, 0.11%, and 0.15%) were fabricated and subjected to crack tip opening displacement (CTOD) testing at −20 °C and −40 °C to evaluate low-temperature HAZ toughness. Results indicate that high-nitrogen TiN steels exhibit superior CTOD values (1.38–2.73 mm) compared to conventional 490-MPa class steels, with no significant reduction in toughness despite increased carbon content. This is attributed to the presence of stable TiN particles, which restrict austenite grain growth during welding thermal cycles, and the formation of fine ferrite–pearlite microstructures in the HAZ. These findings highlight the efficacy of high-nitrogen TiN steels in enhancing low-temperature fracture resistance for welded structures.

1. Introduction

The increasing concentration of populations in urban centers has heightened the demand for land efficiency, leading to a global rise in the construction of super high-rise buildings. Consequently, structural steels used in these buildings—as well as in large-scale offshore structures and ship hulls—are becoming progressively thicker, with ultra-thick steel plates reaching up to 100 mm. This trend toward increased thickness, especially in offshore applications, is accompanied by strict requirements for toughness and strength under low-temperature environments. Multi-layer welding processes for thick plates performed under preheating and interpass temperature maintenance conditions increase both the time and cost of welding, resulting in elevated construction expenses. To mitigate these costs, high heat input welding, which reduces the number of weld passes by utilizing high heat input, is employed to enhance welding productivity and efficiency. Welding, regardless of whether it involves low or high heat input, typically induces a thermal cycle that leads to the coarsening of prior austenite grain size (PAGS) in the heat-affected zone (HAZ), especially near the fusion line. This thermal cycle also results in the formation of various microstructures, including grain boundary ferrite, ferrite side plates, and bainite. The heat input level, however, significantly impacts the extent of this change; a higher heat input promotes a greater degree of PAGS coarsening and a wider HAZ area [1,2]. When welding heat input exceeds 50 kJ/cm, abnormal grain growth occurs during thermal cycles, severely degrading the toughness of the coarse-grained heat-affected zone (CGHAZ) [3,4,5,6,7,8]. These microstructural changes significantly reduce fracture resistance and increase the risk of brittle failure.

To mitigate this, microstructural refinement through the addition of stable precipitates, such as nitrides of Al, Nb, or V, has been explored to restrict austenite grain growth [9,10,11,12,13,14,15,16,17,18,19,20,21]. However, these precipitates (e.g., AlN, Nb(CN), V(CN)) often dissolve at high temperatures near the fusion boundary, thereby limiting their effectiveness [22,23,24]. In contrast, titanium nitride (TiN) particles exhibit superior thermal stability and low solubility, making them highly effective in suppressing austenite grain growth during welding [25,26]. Increasing nitrogen content in TiN-containing steels further reduces TiN solubility in the austenite matrix, thereby enhancing their ability to maintain fine grain structures in the CGHAZ [27,28,29].

Recently, as the significance of fracture mechanics-based evaluation of the HAZ in welded structures has increased, Crack Tip Opening Displacement (CTOD) testing, based on elastic–plastic fracture mechanics, has been widely applied to assess the fracture toughness of the HAZ in various structural steels. In general, steels exhibit a transition from ductile to brittle fracture behavior as the temperature decreases. Since offshore structures are often exposed to extremely low temperatures, they must possess sufficient toughness under such conditions. Therefore, in this study, CTOD testing was conducted to evaluate the low-temperature fracture toughness of the HAZ in high-nitrogen steel in accordance with the ASTM E1290 standard [30]. This study investigates the low-temperature fracture toughness of high-nitrogen TiN-containing steels with varying carbon contents (0.09%, 0.11%, and 0.15%) under different welding heat inputs (45 kJ/cm and 100 kJ/cm). CTOD tests were conducted at −20 °C and −40 °C to evaluate HAZ fracture toughness. Complementary transmission electron microscopy (TEM) analysis was conducted to clarify the relationship between microstructure and toughness.

2. Materials and Methods

2.1. Materials

High-nitrogen steels with a thickness of 40 mm and 490-MPa class strength were produced with carbon contents of 0.09% (Steel A), 0.15% (Steel B), and 0.11% (Steel C). Their chemical compositions and mechanical properties are presented in

Table 1. Chemical compositions of steels used in the experiment (wt.%).

Steel	C	Si	Mn	P	S	Tot- Al	Sol- Al	Ti	B (ppm)	N (ppm)	Ti/N	Ceq 1
A	0.09	0.12	1.48	0.008	0.003	0.050	0.047	0.019	10	130	1.5	0.34
B	0.15	0.12	1.49	0.009	0.003	0.060	0.057	0.017	10	110	1.6	0.40
C	0.11	0.49	1.50	0.008	0.003	0.060	0.057	0.017	10	110	1.5	0.38

¹ Ceq (%) = C + Mn/6 + (Cr + Mo + V)/5 + (Ni + Cu)/15 [IIW].

and

Table 2. Mechanical properties of steels used in the experiment.

Steel	Thickness (mm)	YP (MPa)	TS (MPa)	YR (%)	El. (%)	vE−40 °C (J)
A	40	351	460	76	34	286
B	40	383	518	74	39	235
C	40	405	522	78	27	244

, respectively. The chemical compositions were analyzed using inductively coupled plasma optical emission spectroscopy (ICP-OES) and a carbon-sulfur analyzer (CS). The carbon equivalent (Ceq), which is commonly used to evaluate hardenability and crack susceptibility in the heat-affected zone (HAZ) during welding, was calculated based on the formula recommended by the International Institute of Welding (IIW). The mechanical properties listed in Table 2 were evaluated using a Universal Testing Machine (ZWICK Z600) (ZwickRoell, Ulm, Germany) following the ASTM E8 standard [31]. The Ti/N ratio ranges from 1.5 to 1.6. Charpy impact toughness of the base metal at −40 °C varies between 244 and 286 J, measured at the quarter-thickness position using specimens notched parallel to the rolling direction.

2.2. Welded Specimens

CTOD specimens were prepared using submerged arc welding (ESAB TAF 1250AC, North Bethesda, MD, USA) with heat inputs of 45 kJ/cm for Steels A and B, and 100 kJ/cm for Steel C. A K-bevel groove configuration was adopted to ensure a straight fusion line parallel to the machined notch of the CTOD specimens. The specimens were machined using milling and electrical discharge machining (EDM) to achieve precise geometry and notch placement. Welding with heat inputs of 45 kJ/cm and 100 kJ/cm required eight and four passes, respectively. The base metal was preheated to 80 °C prior to welding, and the interpass temperature was maintained within the range of 150–200 °C, with continuous monitoring using an infrared thermometer.

The welding conditions and parameters are detailed in

Table 3. Applied welding conditions.

Steels	Heat Input	No. of Passes	Groove Angle	Current (A)	Voltage (V)	Welding Speed (cm/min)
A, B	45 kJ/cm	8	60° K-groove	650	36	31
C	100 kJ/cm	4	45° K-groove	900	36	19

, and the groove geometry is illustrated in Figure 1.

2.3. CTOD Testing

Nine CTOD specimens (three per steel type) with dimensions of 38 mm × 76 mm × 550 mm were machined from 40 mm thick welded plates [30]. The welded specimens were machined to a thickness of 38 mm from the 40 mm thick base metal, minimizing material removal to account for weld-induced distortions from multi-layer welding.

To evaluate the fracture toughness of the HAZ, T–L type notches were introduced at distances of 1 mm and 3 mm from the fusion line, as shown in Figure 2. In order to accurately position the notches within the target HAZ region, the fusion line was identified by metallographic etching after machining. This allowed precise placement of notches in the desired microstructural zones. Figure 3 illustrates the sampling orientation of the specimens. Through-thickness notches were machined perpendicular to the weld line direction. After machining the notches, fatigue pre-cracks were introduced and extended by approximately 2.3 mm to achieve a target crack ratio of a/W = 0.53. The total crack length, including the machined notch and the fatigue pre-crack, averaged 40.3 mm.

To mitigate the non-uniform growth of fatigue pre-cracks caused by weld residual stresses, a local compressive load was applied to promote uniform crack propagation through the specimen thickness, in accordance with CTOD test standards. The required compressive load was calculated based on the test specification, and mechanical compression was applied using a rectangular indenter (38 mm × 19 mm) specially fabricated for this purpose.

The upper and lower surfaces of the weld region were compressed by 0.5% of the specimen thickness using a 50-ton dynamic fatigue testing machine, thereby mechanically relieving the residual stresses induced by welding. Fatigue pre-cracks were then introduced, and the specimens were cooled in a liquid nitrogen chamber for over 40 min to reach test temperatures of −20 °C and −40 °C. A clip gauge was installed to measure CTOD, as shown in Figure 4. Standard CTOD specimens with fatigue pre-cracks were immersed in a rectangular chamber filled with 95% alcohol, and liquid nitrogen was introduced to lower the temperature to −20 °C and −40 °C. The specimens were maintained at the target temperature for approximately 40 min to ensure thermal equilibrium. Subsequently, a clip gauge was installed at the notch mouth to measure the CTOD, and the opening displacement was recorded and analyzed. After testing, the fracture surfaces were examined to evaluate the failure modes (ductile or brittle) at the fatigue crack front.

2.4. Microstructural and Fractographic Observations

The microstructures of the base metal and HAZ were examined using optical microscopy (OM) and scanning electron microscopy (SEM). Prior to CTOD testing, metallographic observations were conducted to investigate the microstructural changes in the high-nitrogen steel base metal and welded regions. Specimens were mechanically polished and etched with 4% nital at room temperature for the OM and SEM observations. To investigate the presence and types of precipitates, additional replica specimens were prepared for transmission electron microscopy (TEM) analysis. The replica technique involved a two-step etching process: first, specimens were etched with AA solution (a mixture of acetyl acetate, tetramethylammonium chloride, and methanol); then, precipitates were extracted using a solution of 20% perchloric acid and 80% acetic acid. The extracted films were subsequently analyzed using TEM.

To analyze the morphology and composition of precipitates within the microstructure, TEM was employed using a replica technique. For this purpose, non-magnetic specimens were prepared by etching mechanically mirror-polished surfaces in an ascorbic acid-based solution (890 mL methanol + 100 mL acetylacetone + 10 g tetramethylammonium chloride). After etching, a carbon film was deposited onto the specimen surface. The carbon replica, with precipitates adhered to the film, was then separated in the AA solution and transferred onto copper grids for TEM observation.

3. Results and Discussion

3.1. Low-Temperature CTOD and Microstructure

3.1.1. Heat Input of 45 kJ/cm

CTOD tests conducted at −20 °C and −40 °C for Steels A (0.09% C) and B (0.15% C), both welded with a heat input of 45 kJ/cm, yielded values ranging from 1.38 to 2.73 mm. These results are significantly higher than the typical value of approximately 1 mm reported for conventional 490 MPa-class structural steels [32,33].

Unlike conventional carbon steels, where CTOD values tend to decrease with increasing carbon content, the high-nitrogen TiN steels tested in this study maintained stable CTOD performance regardless of carbon content and test temperature, as shown in Figure 5. Microstructural analysis revealed that Steel B, despite its higher carbon content, exhibited a finer grain structure in both the base metal and heat-affected zone (HAZ) compared to Steel A. Both an increase in carbon content and a finer grain size contribute to greater hardness and strength. However, these two factors can have opposing effects on toughness. In the case of Steel B, the significant grain refinement, a result of the fine precipitates, played a dominant role in enhancing toughness. This positive effect of grain refinement effectively mitigated the potential toughness degradation that would typically be expected from the increased hardness associated with its higher carbon content, leading to an overall improved toughness. The formation of complex TiCN particles, in addition to TiN precipitates in Steel B, is believed to have contributed to this enhanced refinement effect.

For Steel A, CTOD values were similar at 1 mm and 3 mm from the fusion line at both test temperatures. In contrast, Steel B exhibited higher CTOD values at 1 mm compared to 3 mm at −40 °C, despite the coarser microstructure typically expected closer to the fusion line.

Figure 6 shows optical micrographs of the HAZ microstructures observed at 1 mm and 3 mm from the fusion line. The grain size in the CGHAZ of Steel B was finer than that of Steel A. Unlike typical behavior in conventional carbon steels, abnormal grain coarsening in the CGHAZ was not observed. The overall HAZ microstructure primarily consisted of fine polygonal ferrite and pearlite. Although slight coarsening was observed near the fusion boundary, the overall grain size remained comparable to that of the base metal, likely due to the grain growth-inhibiting effect of thermally stable precipitates.

At 1 mm from the fusion line, both steels exhibited mixed microstructures containing fine and coarser polygonal ferrite. At 3 mm, the microstructure transitioned to fine polygonal ferrite and pearlite, closely resembling the base metal. The small differences in CTOD values between Steels A and B are attributed to the similarity in grain sizes within both the fine and coarse regions.

Micro-Vickers hardness measurements were conducted across the weld cross-section from the weld centerline to the base metal using a 500 g load and a dwell time of 12 s. Hardness profiles were measured at three thickness positions—1/4 t, 1/2 t, and 3/4 t—for each specimen. The average hardness values of the base metals for Steels A and B were 160 Hv and 170 Hv, respectively. No evidence of softening was observed in the HAZ for either steel.

3.1.2. Heat Input of 100 kJ/cm

The CTOD test results for Steel C (0.11% C), welded with a heat input of 100 kJ/cm, are shown in Figure 7. The CTOD values were 1.5 mm at −20 °C and 0.3 mm at −40 °C, which are significantly lower than those of the specimens welded at 45 kJ/cm. This reduction in fracture toughness is due to partial dissolution of TiN particles during the high-heat-input welding process, which reduced their effectiveness in inhibiting austenite grain growth. As a result, coarser ferrite grains were formed in the heat-affected zone, leading to reduced low-temperature toughness.

Figure 8 presents the microstructures of Steel C at 1 mm and 3 mm from the fusion line. No significant variation in microstructure was observed between the two locations, and the ferrite grains were larger than those in the 45 kJ/cm specimens. Although high heat input welding typically promotes substantial grain growth and a wider CGHAZ area, Steel C exhibited relatively fine grains and a narrower CGHAZ compared to conventional carbon steels welded under similar heat input conditions. This behavior is presumed to result from grain boundary pinning effects caused by fine precipitates present in the steel. To verify this mechanism, TEM analysis using the replica technique was conducted.

3.1.3. Fracture Observation

Figure 9 and Figure 10 show the SEM fractography of the stable fracture regions in Steels A and B after CTOD testing at −20 °C and −40 °C. Both steels exhibited a mixture of ductile and brittle fracture modes across most specimens, with no significant differences observed with respect to carbon content or distance from the fusion line.

At −20 °C, predominantly ductile fracture features were observed in both steels, while at −40 °C, a mixture of ductile tearing and cleavage fracture was more frequently seen. Overall, these results suggest that plastic deformation occurred near the notch tip at F.L. + 1 mm and that the fracture behavior was not strictly temperature-dependent. Instead, most specimens exhibited a mixed-mode fracture morphology regardless of test temperature or notch location.

3.1.4. TEM Analysis

TEM (JEOL, ARM200) analysis of the HAZ conducted using the replica method revealed that finely dispersed precipitates were present in all specimens.

Selected area diffraction (SAD) patterns confirmed that the nanoscale precipitates, mostly smaller than 10–20 nm, were TiN. These fine TiN particles are believed to have played a significant role in pinning grain boundaries, thereby effectively inhibiting austenite grain growth even under high heat input welding conditions. This grain boundary pinning effect contributed to the maintenance of fine microstructures in the high-nitrogen steels.

However, notable differences in precipitate composition were observed between Steels A and B. As shown in Figure 11, Steel A (0.09% C) primarily contained TiN particles, which were effective in restricting grain growth. In contrast, Steel B (0.15% C) contained both TiN and additional carbide-based precipitates, such as TiC or complex TiCN particles. These additional carbides enhanced grain refinement and contributed to the superior balance of strength and toughness in Steel B. The presence of multiple types of stable precipitates in Steel B explains its stable CTOD values, despite its higher carbon content.

4. Conclusions

This study investigated the influence of carbon content and welding heat input on the fracture toughness of the HAZ in high-nitrogen TiN-containing steels. The key findings of this research lead to the following conclusions:

(1)

High-nitrogen TiN steels with carbon contents of 0.09% and 0.15% exhibited CTOD values ranging from 1.38 to 2.73 mm at −20 °C and −40 °C under a 45 kJ/cm heat input, demonstrating superior fracture toughness compared to conventional 490 MPa-class steels. This improvement is attributed to the uniform distribution of fine TiN precipitates, which effectively restricted austenite grain growth in the HAZ.

(2)

Specimens welded with a high heat input of 100 kJ/cm exhibited reduced CTOD values of 1.5 mm at −20 °C and 0.3 mm at −40 °C. This deterioration in toughness is attributed to the partial dissolution of TiN particles in the CGHAZ, which resulted in insufficient restriction of austenite grain growth.

(3)

Unlike conventional steels, high-nitrogen TiN steels did not show a reduction in CTOD values with increasing carbon content. This stability in toughness is attributed to the formation of TiCN particles in higher-carbon steels, which promoted microstructural refinement and mitigated the adverse effects of carbon.

(4)

Fracture surfaces of the HAZ after CTOD testing at −20 °C and −40 °C exhibited mixed ductile and brittle failure modes, with no significant variation observed across different carbon contents or distances from the fusion line.

These results highlight the potential of high-nitrogen TiN steels to improve the low-temperature toughness of welded offshore structures. However, further investigation is needed to mitigate TiN particle dissolution under high heat input conditions (>100 kJ/cm).