- Email: [email protected]
Effect of titanium addition on the microstructure and inclusion formation in submerged arc welded HSLA pipeline steel
Journal of Materials Processing Technology 209 (2009) 4027–4035
Contents lists available at ScienceDirect
Journal of Materials Processing Technology journal homepage: www.elsevier.com/locate/jmatprotec
Effect of titanium addition on the microstructure and inclusion formation in submerged arc welded HSLA pipeline steel B. Beidokhti ∗ , A.H. Koukabi, A. Dolati Department of Material Science and Engineering, Sharif University of Technology, Azadi Avenue, P.O. Box 11365-9466, Tehran, Iran
a r t i c l e
i n f o
Article history: Received 21 January 2008 Received in revised form 5 August 2008 Accepted 17 September 2008 Keywords: Titanium Acicular ferrite Ductile Inclusion
a b s t r a c t The effect of titanium addition on the SAW weld metal microstructure of API 5L-X70 pipeline steel was investigated. The relationship between microstructure and toughness of the weld deposit was studied by means of full metallographic, longitudinal tensile, Charpy-V notch and HIC tests on the specimens cut transversely to the weld beads. The best combination of microstructure and impact properties was obtained in the range of 0.02–0.05% titanium. By further increasing of titanium content, the microstructure was changed from a mixture of acicular ferrite, grain-boundary ferrite and Widmanstätten ferrite to a mixture of acicular ferrite, grain-boundary ferrite, bainite and ferrite with M/A microconstituent. Therefore, the mode of fracture also changed from dimpled ductile to quasi-cleavage. The results showed an increase in the titanium content of inclusions with increased titanium levels of weld metal. Titanium-base inclusions improve impact toughness by increasing the formation of acicular ferrite in the microstructure. No HIC susceptibility was found in the weld metals with titanium contents less than 0.09%. © 2008 Elsevier B.V. All rights reserved.
1. Introduction The high strength low alloy pipeline steels (HSLA) have a good combination of strength, toughness and weldability. They have been widely used in the construction of long-distance oil and gas transportation systems (Gladman, 1997; Pickering, 1983). Therefore, it is imperative that the steel is characterized by a chemical composition and a microstructure that provide the necessary strength–toughness combination. Generally, the microstructure of conventional C–Mn weld metals consists of varying amounts of acicular ferrite, allotriomorphic ferrite, Widmanstätten ferrite and microphases, with a yield strength ranging from 350 to 450 MPa. Some high strength low alloy weld metals, such as C–Mn with titanium and/or vanadium and niobium additions, exhibit similar microstructures to the C–Mn welds; however, they have higher yield strength, usually in the range of 500–700 MPa (Bose-Filho et al., 2007). It has been reported that a predominant acicular ferrite microstructure with M/A islands as a second phase, exhibits optimum mechanical properties (Contreras et al., 2005; Junhua et al., 2004; Zhao et al., 2002; Zhao and Yang, 2005; Zhong et al., 2006). To meet simultaneously the requirements of both strength and toughness, a class of more heavily alloyed complex steel welding consumables has been developed.
∗ Corresponding author. Tel.: +98 21 66163656; fax: +98 21 66165717. E-mail address: beidokhti [email protected] (B. Beidokhti). 0924-0136/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.jmatprotec.2008.09.021
It has been reported that the weld metal toughness can be increased markedly by an increase of Ni content (Kim et al., 2001). Keehan et al. (2002) found that once Ni exceeds a critical point which depends on Mn percentage; the Charpy toughness at −40 ◦ C is decreased. It was reported by Evans (1998) that the best impact toughness occurred at lower than 0.5 wt% Mo in a controlled manner with respect to Mn content. Also, the addition of Mn and Ni together has been reported to harden weld metal and therefore decrease the impact toughness (Crockett et al., 1995). Conversely, Bhole et al. (2006) found that Mo addition of 0.881 wt% in the weld metal gave the optimal impact toughness at −45 ◦ C with a microstructure of 77% acicular ferrite (AF) and 20% granular bainite (GB). Like many other alloying elements, chromium produces solidsolution strengthening, although the possibility of precipitation hardening should not be ruled out. It was observed that chromium impaired impact toughness even in the beads with large amounts of acicular ferrite (AF). For high Cr contents, AF is replaced by ferrite with second phase (FS) (Jorge et al., 2001). Mo and Nb containing steels are commonly used in pipeline applications because it is observed that the HSLA steels containing Mo and Nb exhibit superior strength and toughness combination as compared to the HSLA steels containing Nb and V (Lee et al., 2000; Sun et al., 2002). Manganese is an important alloying element for solid solution strengthening, however reduced Mn content in steels decreases the centerline microstructural banding (Zhao et al., 2003). Also, manganese can affect the transformation of austenite
4028
B. Beidokhti et al. / Journal of Materials Processing Technology 209 (2009) 4027–4035
Table 1 Chemical composition of the base metal and weld metals. Composition (%)
C
Si
Mn
P
S
Cr
Ni
Mo
V
Ti
Nb
Cu
Al
Base metal Weld no. T00 Weld no. T10 Weld no. T20 Weld no. T30 Weld no. T40 Weld no. T50
0.08 0.06 0.07 0.06 0.06 0.06 0.07
0.23 0.20 0.26 0.26 0.30 0.34 0.35
1.55 1.90 1.92 1.99 2.15 2.23 2.29
0.025 0.030 0.026 0.031 0.029 0.025 0.030
0.011 0.013 0.013 0.015 0.011 0.010 0.015
0.17 0.08 0.08 0.08 0.08 0.08 0.08
0.02 0.02 0.02 0.02 0.02 0.02 0.02
0.01 0.29 0.28 0.28 0.27 0.28 0.26
0.005 0.008 0.01 0.01 0.01 0.01 0.02
0.03 0.004 0.02 0.05 0.09 0.19 0.22
0.03 0.01 0.01 0.01 0.02 0.02 0.02
0.006 0.03 0.03 0.03 0.04 0.05 0.05
0.02 0.003 0.01 0.01 0.02 0.02 0.02
Table 2 The consumable materials for the welding process. Plate thickness (mm)
Electrode type
Electrode diameter (mm)
Flux type
19.8
S2Mo
4
Lincolnweld 995 N
during cooling from high temperatures. Pipeline steels containing low Mn content with additional strength obtained from Cu has been considered for sour service conditions. Inclusions are known to be an important factor in controlling the microstructure and toughness of weld metals, acting as nuclei for acicular ferrite formation and initiation sites for the cleavage fracture process. Bose-Filho et al. (2007) found that in the weld metals with low Ti content, manganese and silicon were the main chemical elements present in inclusions. Increasing the Ti content in the weld metal leads to an increase in the titanium content of inclusions. The methodology in the present study consisted of the application of single pass submerged-arc welding (SAW) to produce the welds with different titanium content. This procedure helps us to investigate the effect of different microstructures and inclusion types due to different titanium contents on the properties of the welds. Simultaneously improvement of these factors is beneficial for the application of HSLA steels in sour environments.
The test method according to NACE Standard TM 0284 (1996) evaluates the resistance of steel to hydrogen induced cracking (HIC). The coupons with 100 mm long × 20 mm wide × wall thickness were cut from the welded plates perpendicular to the weld metal. The test was carried out by immersing specimens in an aqueous solution containing 5% NaCl, 0.5% CH3 COOH saturated with H2 S gas at ambient temperature and pressure. After 96 h exposure in the testing solution, a polished metallographic section of each specimen was inspected for cracks. Three different cracking parameters were measured:
Crack Length Ratio (CLR) =
W
Crack Thickness Ratio (CTR) = Crack Sensitivity Ratio (CSR) =
a
× 100%
T
b
(1)
× 100%
(a × b)
W ×T
× 100%
(2)
(3)
where a is the crack length, b is the crack thickness, W is the section width and T is the test specimen thickness.
3. Results 3.1. Metallographic examination The base metal of all specimens was an API 5L-X70 steel plate contained a typical lamellar ferritic-pearlitic microstructure with
2. Experimental procedure The API 5L-X70 pipeline steel plates were submerged arc welded by using a single pass and similar consumables with different quantities of Fe–Ti powder for designing different chemical compositions of the weld metals. Six groups of specimens with different levels of Ti in the weld metals were made under the same welding conditions. The chemical composition of base metal and the designed contents of Ti in the weld metals are given in Table 1. The consumable materials for the welding process are listed in Table 2. The groove angle and groove depth of the weld joint were 60◦ and 8 mm, respectively. The additive metal powder used was the Fe–Ti powder with approximately 33% Ti and was fed in the weld joint by a custom-made powder-feeder before the welding was started. The electrical data of the welding process are listed in Table 3. The weld was allowed to cool in air with unfused flux on it for 15 min until the temperature dropped below 200 ◦ C. Hardness measurements were made with 10 kg load in a straight line 1 mm below and parallel to the surface of the base metal to cover the complete weld metal, heat affected zone and part of the base metal. For each specimen, 6 full-size standard Charpy impact tests were conducted on six welds at each of two different temperatures, namely −10 and −30 ◦ C. Two longitudinal sub-sized standard tensile round samples (6.0 mm diameter of the reduced section and 30 mm gauge length) consisting completely of the weld metal were used in the tensile tests according to ASTM E8. Optical microscopy and Clemex Imagine Analysis System were used for the microstructural observation and quantitative phase analysis. Also, the microstructural details, type of fracture and composition of inclusions were determined by Tescan Vega II XMU scanning electron microscopy linked to a Rontec EDS system.
Table 3 The electrical parameters of the welding process. Voltage (V)
Current (A)
Welding speed (cm/min)
Stick out (mm)
30
400
25.4
28
Fig. 1. SEM micrographs of the HAZ microstructure: (a) coarse grained zone; (b) fine grained zone.
B. Beidokhti et al. / Journal of Materials Processing Technology 209 (2009) 4027–4035
4029
Table 4 Quantitative microstructural analysis. Weld number
Acicular ferrite (AF)%
Grain boundary Widmannstätten ferrite (GBF)% ferrite (WF)%
Other phases (B and ferrite with M/A)%
T00 T10 T20 T30 T40 T50
70.1 74.6 71.7 61.0 54.2 48.1
15.8 13.6 13.5 14.7 13.0 15.3
5.9 6.5 8.9 19.7 27.9 32.8
8.2 5.3 5.9 4.6 4.9 3.8
Fig. 3. The variation of: (a) Impact energy; (b) shear fraction area in different titanium contents of the weld metals.
Fig. 2. Hardness values vs. titanium percentage of the weld metals.
13.7% pearlite. The application of single pass welding and same base material created a similar heat affected zone (HAZ) in all welded specimens. The microstructures of HAZ were a mixture of ferrite and bainite in the coarse grained zone and equiaxed ferrite, pearlite
Fig. 4. Microstructure of the weld metals with different titanium contents: (a) T00, a mixture of acicular and grain boundary ferrite; (b and c) T10 and T20, increasing of acicular ferrite in the microstructure; (d) T30, bainitic region in the microstructure.
4030
B. Beidokhti et al. / Journal of Materials Processing Technology 209 (2009) 4027–4035
Table 5 Tensile properties. Weld number
Yield strength (MPa)
Tensile strength (MPa)
Elongation (%)
T00 T10 T20 T30 T40 T50
570 597 612 647 768 795
659 681 758 809 933 947
21.2 22.7 20.4 19.8 19.0 17.7
Table 6 Comparison of the inclusion chemical composition in different weld metals. Weld number
Ti (%)
Mn (%)
Si (%)
Al (%)
T00 T10 T20 T30 T40 T50
6.0 21.2 25.9 41.7 58.4 62.8
32.7 24.7 19.2 12.2 14.3 17.1
14.2 8.1 7.2 6.3 6.0 4.6
10.2 16.4 15.6 20.9 13.7 12.4
and a few M/A microconstituents in the fine grained zone, respectively (Fig. 1). The change of each microstructure in the weld metals with different level of alloying elements is shown in Table 4. For all weld metals, evaluations were performed to determine the amounts of acicular ferrite (AF), grain-boundary ferrite (GBF), Widmanstätten ferrite (WF) and other phases like bainite (B) and M/A microconstituents within the weld metal. 3.2. Mechanical properties The hardness values were increased by addition of titanium to the weld metal as shown in Fig. 2. The yield strength, tensile strength and percentage of elongation data obtained in the longitudinal direction of the welds are summarized in Table 5. The longitudinal yield strength and tensile strength were in the range of 570–795 and 659–947 MPa, respectively. The variation of Charpy-V notch impact energy and the percentage of shear fraction area of the weld metals at two degrees, −10 and −30 ◦ C, are shown in Fig. 3. It is observed that the best impact properties were attained for 0.02–0.05% Ti in both cases. Fig. 5. SEM micrographs of the weld metals: (a) T00, acicular ferrite and grain boundary ferrite; (b) T10, rich acicular ferrite microstructure; (c) T50, formation of bainite.
3.3. Inclusion analysis Table 6 lists the variations of inclusion chemical composition for different weld metals. According to Bose-Filho et al. (2007) findings, increasing the titanium content of weld metal promoted the titanium levels of inclusions.
high in the specimens T40 and T50. Weld no. T30 with 0.09% Ti (and low notch toughness) showed the HIC susceptibility as same as the lower titanium and manganese content weld metals.
3.4. HIC test
4. Discussion
The values of CLR, CTR and CSR for each weld are presented in Table 7. According to the results, the cracking parameters were only
Comparison of the weld compositions showed that increased titanium content promoted the amounts of manganese and silicon in the weld metal. The basic-alumina flux has a basicity index of about 1.3. This flux produces about 0.5% Mn and 0.2% Si transfer to the weld deposit. With addition of Ti to the flux, the recovery of Mn and Si was increased and the amounts of these elements in the weld metal were promoted. In the welds with high contents of Ti, the difference of manganese percentage between the weld metal and base metal was about 0.7%. Therefore, in titanium additions greater than 0.09%, the positive effect of titanium is in contest with the negative effect due to the high manganese content of weld metal.
Table 7 The cracking parameters of HIC test for each weld metal. Weld number
CLR (%)
CTR (%)
CSR (%)
T00 T10 T20 T30 T40 T50
10.7 8.5 12.3 14.4 24.5 28.4
3.1 1.5 4.4 4.2 3.5 4.7
0.3 0.2 0.8 1.3 2.8 3.5
B. Beidokhti et al. / Journal of Materials Processing Technology 209 (2009) 4027–4035
4031
Fig. 6. Fracture mode of specimens at −10 ◦ C: (a) T00; (b) T10; (c) T30; (d) T40.
It can be seen from the corresponding micrographs shown in Fig. 4 that a mixed microstructure with acicular ferrite (AF), grain-boundary ferrite (GBF) and Widmanstätten ferrite (WF) was obtained in each specimen. In the welds with 0.09% Ti and more, bainite and ferrite with secondary phase (M/A) were also found in the microstructure. This phenomenon is related to the high amount of manganese recovery. Also according to Evans (1986, 1992) for the titanium content greater than 30 ppm, the weld metal contains M/A constituents as a microphase. In the specimens with titanium content up to 0.05%, formation of acicular ferrite was promoted. Therefore, due to a competitive growth of acicular ferrite and Widmanstätten ferrite, the amount of Widmanstätten ferrite was decreased. Spontaneously, M/A constituents were promoted because the amount of alloying elements and hardenability were increased. The results of hardness evaluation correlated to the microstructure observations. When titanium was added to the weld metal, hardness values were increased; but up to 0.05% Ti, these values were lower than the acceptance criterion (248 HV) for sour service in accordance with NACE MR0175. The titanium addition in the range of 0.02–0.05% in the weld metal gave the optimal hardness value and microstructure. Increased titanium content of the weld metal greater than 0.09%, leads to an increase in the amount of bainite and M/A microconstituents and promotes hardness values. In the specimen contained 0.22% Ti, the hardness value was about
321 HV. This hardness is larger than NACE restriction and therefore the risk of cracking in sour service is increased. Previous researchers (Bose-Filho et al., 2007) evaluated the hardness of bainite and low carbon martensite in the weld metal of HSLA pipeline steel to be approximately 325 HV and 340 HV, respectively. The weld metal contained 0.09% Ti exhibited the hardness value of about 286 HV; therefore, formation of martensite phase in this specimen was less pronounced. On the other hand, if the flux with basicity index of greater than 2 is used, the manganese content of weld metal will be reduced. Then, it is predicted that the optimal properties can be obtained in as-deposited welds with higher Ti content, e.g. at 0.09% Ti. Manganese is an important alloying element for solid solution strengthening; also, manganese can affect the transformation of austenite during cooling from high temperatures. It is observed that addition of titanium increases manganese values in the weld metals. Increased manganese content promoted formation of hard phases, especially bainite in the microstructure of weld metals with Ti ≥ 0.09%. Fig. 5 shows SEM micrographs of the different weld metal microstructures. The nucleation of acicular ferrite was increased by addition of titanium (Fig. 5a and b); beyond the optimal range of titanium (0.02–0.05%), increased manganese of the weld metal was a dominant factor to determining the weld metal microstructure. Therefore, bainite was observed in the large areas of the weld metal T50 (Fig. 5c).
4032
B. Beidokhti et al. / Journal of Materials Processing Technology 209 (2009) 4027–4035
Fig. 7. Fracture mode of specimens at −30 ◦ C: (a) T00; (b) T10; (c) T30; (d) T40. The quasi-cleavage fracture extended in the larger regions of all specimens in comparison to −10 ◦ C ones.
Also, increased titanium content leads to an increase in the yield and also in the tensile stresses. In the low titanium weld metals, the elongation percentage was promoted; but when B and M/A appeared in the microstructure, this parameter was dropped. Decreased elongation of the 0.05% Ti weld metal, with a good distribution of acicular ferrite, was only 0.8% in comparison to the 0.004% Ti weld metal. The results of tensile test exhibited that Ti addition had no strong negative effect on the elongation values; e.g. the weld metal with titanium content of about 0.19% (and consequently 2.23% Mn) exhibited the elongation value of about 19%. Generally, the strengthening and improvement of HIC resistance for this AF pipeline steel were obtained by addition of titanium. In submerged arc deposits, the fine oxide inclusion dispersion might act as sites for TiN and TiC precipitation. The precipitated titanium carbonitrides are expected to provide dispersion strengthening and to act as beneficial hydrogen traps in hydrogen or hydrogen sulfide environments. However, the yield strength and tensile strength of the weld metal were increased by further addition of alloying elements and also the elongation values were acceptable; but the negative effect due to the formation of B and M/A appeared in the Charpy impact test. The Charpy impact results confirmed previous findings. The addition of Ti up to 0.05% increased the impact energy value and the percentage of shear fracture area. Then, according to the forma-
tion of bainite and ferrite with M/A, these factors were decreased. At −10 ◦ C, the highest impact energy of about 137 J was obtained in the weld metal with 0.02% titanium. Also, the 0.05% Ti weld metal gave the impact properties superior to the reference specimen (weld no. T00). In the higher titanium content specimens, impact energy was decreased and reached to 28 J for the weld metal T50 with 0.22% Ti–2.29% Mn. Scanning electron microscopy analysis of the fractured impact specimens were undertaken to characterize the mode of fracture in the weld metals. Figs. 6 and 7 show the fractured surface micrographs with the wide ranges of fracture mode at −10 and −30 ◦ C, respectively. The evolution of the fracture mechanism was evident from dimpled ductile to quasi-cleavage mode with increased the amounts of titanium and manganese. This condition revealed the high degree of hydrogen embrittlement risk in sour environments (the absorbed energy of about 23 J for the 0.22% Ti specimen at −30 ◦ C). The impact energies and the shear fracture areas were impressively decreased at −30 ◦ C. In the weld metals with titanium contents of about 0.02 and 0.05%, the impact energy values were 116 and 101 J, respectively; whereas, this value for the reference specimen was only 78 J. These findings imply that the positive effect of Ti on the impact properties is increased with decreased temperature. Increasing the amount of acicular ferrite due to titanium addition could be responsible for this improvement. The formation of hard
B. Beidokhti et al. / Journal of Materials Processing Technology 209 (2009) 4027–4035
4033
phases like B and M/A due to increased manganese of the high alloy weld metals, leads to impairment of toughness again. The interesting result was the lack of any great difference between impact energies of the weld metals with titanium additions in the range of 0.09–0.22%. Increased titanium content in this range leads to progressive increasing of manganese values in the weld metals. It seems under this condition, manganese is the controlling parameter of impact properties. Because the amount of manganese in the specimens are high enough and the variations of manganese content in this range are negligible, these weld metals have almost the same impact toughness. Decreased impact energies in the high titanium content weld metals could be the result of two conditions: 1. Increasing the manganese content in the weld metal resulted from the recovery of manganese. 2. The strengthening effect of titanium due to the excessive distribution of fine Ti-base particles in the microstructure. For attaining different titanium contents of the weld metals without impressive change in the manganese content, welding with high basicity index flux is recommended. The non-metallic inclusions observed in this work are the result of deoxidation and/or desulphurization reactions taking place in the weld pool. It has been shown by some researchers (Abson et al., 1978; Cochrane and Kirkwood, 1978; Ricks et al., 1982) that these inclusions are very important in the microstructural development of ferritic weld metals since they act as nuclei for acicular ferrite. It has also been suggested that inclusions may pin austenite grain boundaries and reduce their grain size. These inclusions may also be responsible for determining the fracture resistance of HSLA steel weld metals (Bose-Filho, 1995). Garbarz et al. (2003) investigated on the type of sulfide inclusions in steels. Inclusions were identified by microanalysis in thin foils as MnS, (Mn, Cu)S, CuS and Cu2 S. Bose-Filho et al. (2007) showed that inclusions of the weld metal with a classical microstructure were mainly MnO or/and MnS and SiO2 for a low titanium content. Increasing the titanium content of the weld led to an increase in titanium within the inclusions, and at 230 ppm or more of titanium in the weld, Ti2 O3 and MnO and/or MnS were the main inclusion constituents. In the weld metals with very high titanium content, 700 ppm, the amount of titanium in the inclusions varied in the range of 60–70%. Also Yuan (2007) found that Ti-stabilized ultra low carbon steels contained a large number of fine Ti(C, N) and coarse Ti4 C2 S2 precipitates. According to Bose-Filho et al. (2007), the results of our paper showed an increase in the titanium content of inclusions with increased titanium content of the weld metal. This phenomenon leads to change in the nature of inclusions and decreases the amount of Mn-rich inclusions such as MnS. Bose-Filho et al. (2007) examined this phenomenon at the lower percentage of titanium addition and higher alloying elements. Also, they did not determine the optimum percentage of titanium. The Ti-base inclusions improve impact toughness by increasing the formation of acicular ferrite in the microstructure. North et al. (1979) showed that titanium and vanadium promote the acicular ferrite formation in the CMnCb deposits, but aluminum, zirconium and rare earth metals do not. The principal driving forces for growth of a ferrite crystal from austenite comprise the difference in chemical free energy between ferrite and austenite, and the tendency for the system to minimize the total interfacial free energy of interface boundaries (Aronson, 1960). In the case of TiC and TiN, close matching between the (1 0 0) planes of ferrite and these particles occurs and the mismatch parameter (ı) is very small. For ı < 0.05 the interface is coherent, and for 0.05 < ı < 0.25 a semi-coherent boundary
Fig. 8. (a) HIC crack in the mid-thickness of base metal; (b) SOHIC crack in the bainitic weld metal.
is produced (Chadwick, 1972). Since coherent interface has a much less interfacial energy than semi-coherent interface, therefore TiC and TiN are favorable sites for ferrite lath nucleation. According to these findings, the results of our work showed that the addition of titanium up to 0.05% increased the formation of acicular ferrite in the microstructure; but increased manganese amounts in the high titanium content weld metals hindered development of acicular ferrite and encouraged formation of bainite and ferrite with M/A in the microstructure (Table 4). Also, several authors (Garet et al., 1998; Lee and Lee, 1984, 1987) have shown that MnS and TiC inclusions are temporary and permanent traps for hydrogen atoms, respectively. Consequently, the substitution of manganese inclusion with titanium one promotes the number of permanent trap sites for hydrogen. It is predicted that this fact improves the resistance of steel in the sour environment. The HIC investigation was done to determine the crack susceptibility of different specimens in the hydrogen containing environments. The microscopic studies showed that the HIC cracks were propagated significantly in the base metal; only the specimens T40 and T50 exhibited cracking in the weld metal (Fig. 8). These specimens had a very high hardness and low Charpy toughness as result of the high percentages of titanium and manganese in the weld metal. No crack was observed in the weld metal T30 (0.09% Ti) contained a relatively high percentage of bainite and M/A microconstituents. As was mentioned before, if the flux with basicity index of greater than 2 is used, the manganese content of weld metal will be reduced. Then, it is predicted that the good impact toughness and HIC resistance can be obtained in the as-deposited welds with higher Ti content (the 0.09% Ti weld metal, coincidentally, showed low toughness and no HIC susceptibility). The CLR value less than 15% was adopted as an acceptance criterion of HIC test according to Kushida et al. (1997). On the basis of the CLR values, only the weld metals T40 and T50 are not acceptable.
4034
B. Beidokhti et al. / Journal of Materials Processing Technology 209 (2009) 4027–4035
All specimens exhibited the same CLR, CTR and CSR values; but, the weld metal cracking increased these values in the specimens T40 and T50. In the base metal, the HIC cracks were observed along the segregation band in the mid-thickness as shown in Fig. 8a. The morphology of the weld metal cracks was different. According to previous findings (Carneiro et al., 2003; Kane and Cayard, 1998), the short cracks normally associated with HIC can form a stacked array in the through thickness direction under the influence of residual tensile stress. This phenomenon is referred to as stress oriented hydrogen induced cracking (SOHIC). In the specimens T40 and T50, the SOHIC cracks were seen in the weld metals containing high amount of bainite (Fig. 8b). These results revealed no HIC susceptibility in the weld metals with titanium contents less than 0.09%; but, it is impossible to find which one has the highest HIC resistance. Furthermore, all specimens in current investigation have high yield strength (≥535 MPa); in conformity with NACE TM0177 for sour service application, sulfide stress cracking (SSC) studies will be requiring in the future studies.
in all weld metals, except in the weld metals with high titanium content as much as 0.19%. The SOHIC type cracks were formed in the bainitic microstructure of these welds. 7. MnS and TiC inclusions are temporary and permanent traps for hydrogen atoms, respectively. The substitution of manganese inclusion with titanium one, promotes the number of permanent trap sites for hydrogen and improves the resistance of steel in sour environments. Acknowledgements The authors are thankful to National Iranian Gas Company (NIGC), for financial supporting of this project and SAFA Co., Saveh, for providing the materials and laboratory facility in their factory. Thanks are also due to A. Erfanfar, R. Shamloo, A. Zandipoor and A. Khoshakhlagh of SAFA Co., for their assistance and critical review of the work and J. Woolington and N. Beidokhti for assistance with proofreading. References
5. Conclusions The effect of titanium addition on the microstructure and toughness of weld metals in an API 5L-X70 pipeline steel produced by submerged arc welding technique has been investigated. Based on the obtained results, the conclusions are as follows: 1. With addition of Ti to the flux, the recovery of Mn and Si were increased and the amounts of these elements in the weld metal were promoted. In the specimens with high titanium contents, the difference of manganese percentage between weld metal and base metal was about 0.7%. Therefore, in the titanium addition greater than 0.09%, the positive effect of Ti is in contest with the negative effect due to the high Mn content of the weld metal. 2. As a result of increasing the titanium content, the amount of acicular ferrite was increased in the weld metal. With further addition of titanium, the microstructure has changed from a mixture of acicular ferrite, grain-boundary ferrite and Widmanstätten ferrite to a mixture of acicular ferrite, grainboundary ferrite, bainite and ferrite with M/A microconstituents. The recovery of manganese and the hardening effect of this element in the high titanium content weld metals, encouraged formation of hard phases like B and M/A in the weld microstructure. 3. The addition of Ti up to 0.05% increased the impact energy value and the percentage of shear fracture area. Afterward, according to the formation of harder phases such as B and M/A, these factors were decreased. The evolution of the fracture mechanism was evident from dimpled ductile to quasi-cleavage mode with increasing the amount of titanium and manganese. The positive effect of Ti on the impact properties increased with decreased temperature. 4. The best combination of microstructure and impact properties was obtained in the range of 0.02–0.05% Ti. The hardness values of these specimens were lower than the acceptance criterion (248 HV) for sour services in accordance to NACE MR0175. 5. The results showed an increase in the titanium content of inclusions with increasing the titanium content of the weld metal. This phenomenon causes a change in the nature of inclusions and decreases the amount of Mn-rich inclusions such as MnS. The Ti-base inclusions improve impact toughness by increasing the formation of acicular ferrite in the microstructure. 6. In the base metal, the HIC cracks were observed along the segregation band in the mid-thickness. No HIC crack was observed
Abson, D.J., Dolby, R.E., Hart, P.H.M., 1978. The role of nonmetallic inclusions in ferrite nucleation in carbon steel weld metals. In: International Conference on Trends in Steels and Consumables for Welding, London, UK, pp. 75–89. Aronson, H.J., 1960. The proeutectoid ferrite and proeutectoid cementite reactions. In: Met. Soc. AIME Symposium on Decomposition of Austenite by Diffusional Processes, Philadelphia, pp. 387–546. Bhole, S.D., Nemade, J.B., Collins, L., Liu, C., 2006. Effect of nickel and molybdenum additions on weld metal toughness in a submerged arc welded HSLA line-pipe steel. J. Mater. Process. Technol. 173, 92–100. Bose-Filho, W.W., 1995. Micromechanisms of brittle fracture in high strength low alloy steel weld metals. Ph.D. Thesis. The University of Birmingham, Birmingham, UK. Bose-Filho, W.W., Carvalho, A.L.M., Strangwood, M., 2007. Effect of alloying elements on the microstructure and inclusion formation in HSLA multipass welds. Mater. Charact. 58 (1), 29–39. Carneiro, R.A., Ratnapuli, R.C., Lins, V.F.C., 2003. The influence of chemical composition and microstructure of API linepipe steels on hydrogen induced cracking and sulfide stress corrosion cracking. Mater. Sci. Eng. A 357, 104–110. Chadwick, G.A., 1972. Metallography of Phase Transformations. Butterworth Publications, London, pp. 193–200. Cochrane, R.C., Kirkwood, P.R., 1978. The effects of oxygen on weld metal microstructure. In: The Welding Institute Conference on Trends in Steels and Consumables, London, UK, pp. 103–122. Contreras, A., Albiter, A., Salazar, M., Perez, R., 2005. Slow strain rate corrosion and fracture characteristics of X-52 and X-70 pipeline steels. Mater. Sci. Eng. A 407, 45–52. Crockett, D.D., Rhone, J.A., Young, R.F., Noernberg, D.C., 1995. Design considerations for submerged arc consumables intended for the manufacture of line pipe. Pipeline Technol. 1, 151–162. Evans, G.M., 1986. The effect of silicon on the microstructure and properties of C–Mn all-weld metal deposits. Met. Constr. 28, 438R–444R. Evans, G.M., 1992. Effect of micro-alloying elements in C–Mn steel weld metal. In: ADESOL Conference, Madrid, Spain. Evans, G.M., 1998. Effect of molybdenum on microstructure and properties of C–Mn all-weld metal deposits. Join. Mater. 1 (5), 239–246. Garbarz, B., Marcisz, J., Wojtas, J., 2003. TEM analysis of fine sulphides dissolution and precipitation in steel. Mater. Chem. Phys. 81, 486–489. Garet, M., Brass, A.M., Haut, C., Guttierez-Solana, F., 1998. Hydrogen trapping on non metallic inclusions in Cr–Mo low alloy steels. Corros. Sci. 40, 1073–1086. Gladman, T., 1997. The development of microalloyed steels. In: The Physical Metallurgy of Microalloyed Steels, The Institute of Materials, London, England, pp. 1–18. Jorge, J.C.F., Souza, L.F.G., Rebello, J.M.A., 2001. The effect of chromium on the microstructure/toughness relationship of C–Mn weld metal deposits. Mater. Charact. 47, 195–205. Junhua, K., Lin, Z., Bin, G., Pinghe, L., Aihua, W., Changsheng, X., 2004. Influence of Mo content on microstructure and mechanical properties of high strength pipeline steel. Mater. Des. 25, 723–728. Kane, R.D., Cayard, M.S., 1998. Roles of H2 S in the behavior of engineering alloys: a review of literature and experience. In: NACE, Corrosion/98. Paper No. 274. Houston, USA. Keehan, E., Andren, H.O., Karlsson, L., Murugananth, M., Bhadeshia, H.K.D.H., 2002. Microstructural and mechanical effects of nickel and manganese on high strength steel weld metals. In: Sixth International Conference: Trend in Welding Research, Pine Mountain, GA, USA, pp. 695–700. Kim, S.H., Bang, C.Y., Bang, K.S., 2001. Weld metal impact toughness of electron beam welded 9% Ni steel. J. Mater. Sci. 36, 1197–2000.
B. Beidokhti et al. / Journal of Materials Processing Technology 209 (2009) 4027–4035 Kushida, T., Okaguchi, S., Hamada, M., Yamamoto, A., Ohnishi, K., Fujino, J., 1997. Study of X80 grade high strength line pipe for sour service. In: NACE, Corrosion/97. Paper No. 24. Houston, USA. Lee, H.G., Lee, J.Y., 1984. Hydrogen trapping by TiC particles in iron. Acta Metall. 32, 131–136. Lee, J.Y., Lee, J.L., 1987. A trapping theory of hydrogen in pure iron. Philos. Mag. A 56, 293–309. Lee, W.B., Hong, S.G., Park, C.G., Kim, K.H., Park, S.H., 2000. Influence of Mo on precipitation hardening in hot rolled HSLA steels containing Nb. Scripta Mater. 43, 319–324. North, T.H., Bell, H.B., Koukabi, A., Craig, I., 1979. Notch toughness of low oxygen content submerged arc deposits. Weld. J. 58 (12), 343s–354s. Pickering, F.B., 1983. The spectrum of microalloyed high strength low alloy steels. In: HSLA Steels Technology and Applications, Conference Proceedings of International Conference on Technology and Applications of HSLA Steels, pp. 1–32. Ricks, R.A., Barrite, G.S., Howell, P.R., 1982. The nature of acicular ferrite in HSLA steel weld metals. J. Mater. Sci. 17, 732–740.
4035
Sun, W., Lu, C., Tieu, A.K., Jiang, Z., Liu, X., Wang, G., 2002. Influence of Nb, V and Ti on peak strain of deformed austenite in Mo-based micro-alloyed steels. Mater. Process. Technol. 125/126, 72–76. Yuan, X., 2007. Precipitates and hydrogen permeation behavior in ultra-low carbon steel. Mater. Sci. Eng. A 452/453, 116–120. Zhao, M.C., Yang, K., 2005. Strengthening and improvement of sulfide stress cracking resistance in acicular ferrite pipeline steels by nano-sized carbonitrides. Scripta Mater. 52, 881–886. Zhao, M.C., Yang, K., Shan, Y., 2002. The effects of thermo-mechanical control process on microstructures and mechanical properties of a commercial pipeline steel. Mater. Sci. Eng. A 335, 14–20. Zhao, M.C., Yang, K., Shan, Y., 2003. Continuous cooling transformation of undeformed and deformed low carbon pipeline steels. Mater. Sci. Eng. A 335, 126– 136. Zhong, Y., Xiao, F., Zhang, J., Shan, Y., Wang, W., Yang, K., 2006. In situ TEM study of the effect of M/A films at grain boundaries on crack propagation in an ultra-fine acicular ferrite pipeline steel. Acta Mater. 54, 435–443.
Contents lists available at ScienceDirect
Journal of Materials Processing Technology journal homepage: www.elsevier.com/locate/jmatprotec
Effect of titanium addition on the microstructure and inclusion formation in submerged arc welded HSLA pipeline steel B. Beidokhti ∗ , A.H. Koukabi, A. Dolati Department of Material Science and Engineering, Sharif University of Technology, Azadi Avenue, P.O. Box 11365-9466, Tehran, Iran
a r t i c l e
i n f o
Article history: Received 21 January 2008 Received in revised form 5 August 2008 Accepted 17 September 2008 Keywords: Titanium Acicular ferrite Ductile Inclusion
a b s t r a c t The effect of titanium addition on the SAW weld metal microstructure of API 5L-X70 pipeline steel was investigated. The relationship between microstructure and toughness of the weld deposit was studied by means of full metallographic, longitudinal tensile, Charpy-V notch and HIC tests on the specimens cut transversely to the weld beads. The best combination of microstructure and impact properties was obtained in the range of 0.02–0.05% titanium. By further increasing of titanium content, the microstructure was changed from a mixture of acicular ferrite, grain-boundary ferrite and Widmanstätten ferrite to a mixture of acicular ferrite, grain-boundary ferrite, bainite and ferrite with M/A microconstituent. Therefore, the mode of fracture also changed from dimpled ductile to quasi-cleavage. The results showed an increase in the titanium content of inclusions with increased titanium levels of weld metal. Titanium-base inclusions improve impact toughness by increasing the formation of acicular ferrite in the microstructure. No HIC susceptibility was found in the weld metals with titanium contents less than 0.09%. © 2008 Elsevier B.V. All rights reserved.
1. Introduction The high strength low alloy pipeline steels (HSLA) have a good combination of strength, toughness and weldability. They have been widely used in the construction of long-distance oil and gas transportation systems (Gladman, 1997; Pickering, 1983). Therefore, it is imperative that the steel is characterized by a chemical composition and a microstructure that provide the necessary strength–toughness combination. Generally, the microstructure of conventional C–Mn weld metals consists of varying amounts of acicular ferrite, allotriomorphic ferrite, Widmanstätten ferrite and microphases, with a yield strength ranging from 350 to 450 MPa. Some high strength low alloy weld metals, such as C–Mn with titanium and/or vanadium and niobium additions, exhibit similar microstructures to the C–Mn welds; however, they have higher yield strength, usually in the range of 500–700 MPa (Bose-Filho et al., 2007). It has been reported that a predominant acicular ferrite microstructure with M/A islands as a second phase, exhibits optimum mechanical properties (Contreras et al., 2005; Junhua et al., 2004; Zhao et al., 2002; Zhao and Yang, 2005; Zhong et al., 2006). To meet simultaneously the requirements of both strength and toughness, a class of more heavily alloyed complex steel welding consumables has been developed.
∗ Corresponding author. Tel.: +98 21 66163656; fax: +98 21 66165717. E-mail address: beidokhti [email protected] (B. Beidokhti). 0924-0136/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.jmatprotec.2008.09.021
It has been reported that the weld metal toughness can be increased markedly by an increase of Ni content (Kim et al., 2001). Keehan et al. (2002) found that once Ni exceeds a critical point which depends on Mn percentage; the Charpy toughness at −40 ◦ C is decreased. It was reported by Evans (1998) that the best impact toughness occurred at lower than 0.5 wt% Mo in a controlled manner with respect to Mn content. Also, the addition of Mn and Ni together has been reported to harden weld metal and therefore decrease the impact toughness (Crockett et al., 1995). Conversely, Bhole et al. (2006) found that Mo addition of 0.881 wt% in the weld metal gave the optimal impact toughness at −45 ◦ C with a microstructure of 77% acicular ferrite (AF) and 20% granular bainite (GB). Like many other alloying elements, chromium produces solidsolution strengthening, although the possibility of precipitation hardening should not be ruled out. It was observed that chromium impaired impact toughness even in the beads with large amounts of acicular ferrite (AF). For high Cr contents, AF is replaced by ferrite with second phase (FS) (Jorge et al., 2001). Mo and Nb containing steels are commonly used in pipeline applications because it is observed that the HSLA steels containing Mo and Nb exhibit superior strength and toughness combination as compared to the HSLA steels containing Nb and V (Lee et al., 2000; Sun et al., 2002). Manganese is an important alloying element for solid solution strengthening, however reduced Mn content in steels decreases the centerline microstructural banding (Zhao et al., 2003). Also, manganese can affect the transformation of austenite
4028
B. Beidokhti et al. / Journal of Materials Processing Technology 209 (2009) 4027–4035
Table 1 Chemical composition of the base metal and weld metals. Composition (%)
C
Si
Mn
P
S
Cr
Ni
Mo
V
Ti
Nb
Cu
Al
Base metal Weld no. T00 Weld no. T10 Weld no. T20 Weld no. T30 Weld no. T40 Weld no. T50
0.08 0.06 0.07 0.06 0.06 0.06 0.07
0.23 0.20 0.26 0.26 0.30 0.34 0.35
1.55 1.90 1.92 1.99 2.15 2.23 2.29
0.025 0.030 0.026 0.031 0.029 0.025 0.030
0.011 0.013 0.013 0.015 0.011 0.010 0.015
0.17 0.08 0.08 0.08 0.08 0.08 0.08
0.02 0.02 0.02 0.02 0.02 0.02 0.02
0.01 0.29 0.28 0.28 0.27 0.28 0.26
0.005 0.008 0.01 0.01 0.01 0.01 0.02
0.03 0.004 0.02 0.05 0.09 0.19 0.22
0.03 0.01 0.01 0.01 0.02 0.02 0.02
0.006 0.03 0.03 0.03 0.04 0.05 0.05
0.02 0.003 0.01 0.01 0.02 0.02 0.02
Table 2 The consumable materials for the welding process. Plate thickness (mm)
Electrode type
Electrode diameter (mm)
Flux type
19.8
S2Mo
4
Lincolnweld 995 N
during cooling from high temperatures. Pipeline steels containing low Mn content with additional strength obtained from Cu has been considered for sour service conditions. Inclusions are known to be an important factor in controlling the microstructure and toughness of weld metals, acting as nuclei for acicular ferrite formation and initiation sites for the cleavage fracture process. Bose-Filho et al. (2007) found that in the weld metals with low Ti content, manganese and silicon were the main chemical elements present in inclusions. Increasing the Ti content in the weld metal leads to an increase in the titanium content of inclusions. The methodology in the present study consisted of the application of single pass submerged-arc welding (SAW) to produce the welds with different titanium content. This procedure helps us to investigate the effect of different microstructures and inclusion types due to different titanium contents on the properties of the welds. Simultaneously improvement of these factors is beneficial for the application of HSLA steels in sour environments.
The test method according to NACE Standard TM 0284 (1996) evaluates the resistance of steel to hydrogen induced cracking (HIC). The coupons with 100 mm long × 20 mm wide × wall thickness were cut from the welded plates perpendicular to the weld metal. The test was carried out by immersing specimens in an aqueous solution containing 5% NaCl, 0.5% CH3 COOH saturated with H2 S gas at ambient temperature and pressure. After 96 h exposure in the testing solution, a polished metallographic section of each specimen was inspected for cracks. Three different cracking parameters were measured:
Crack Length Ratio (CLR) =
W
Crack Thickness Ratio (CTR) = Crack Sensitivity Ratio (CSR) =
a
× 100%
T
b
(1)
× 100%
(a × b)
W ×T
× 100%
(2)
(3)
where a is the crack length, b is the crack thickness, W is the section width and T is the test specimen thickness.
3. Results 3.1. Metallographic examination The base metal of all specimens was an API 5L-X70 steel plate contained a typical lamellar ferritic-pearlitic microstructure with
2. Experimental procedure The API 5L-X70 pipeline steel plates were submerged arc welded by using a single pass and similar consumables with different quantities of Fe–Ti powder for designing different chemical compositions of the weld metals. Six groups of specimens with different levels of Ti in the weld metals were made under the same welding conditions. The chemical composition of base metal and the designed contents of Ti in the weld metals are given in Table 1. The consumable materials for the welding process are listed in Table 2. The groove angle and groove depth of the weld joint were 60◦ and 8 mm, respectively. The additive metal powder used was the Fe–Ti powder with approximately 33% Ti and was fed in the weld joint by a custom-made powder-feeder before the welding was started. The electrical data of the welding process are listed in Table 3. The weld was allowed to cool in air with unfused flux on it for 15 min until the temperature dropped below 200 ◦ C. Hardness measurements were made with 10 kg load in a straight line 1 mm below and parallel to the surface of the base metal to cover the complete weld metal, heat affected zone and part of the base metal. For each specimen, 6 full-size standard Charpy impact tests were conducted on six welds at each of two different temperatures, namely −10 and −30 ◦ C. Two longitudinal sub-sized standard tensile round samples (6.0 mm diameter of the reduced section and 30 mm gauge length) consisting completely of the weld metal were used in the tensile tests according to ASTM E8. Optical microscopy and Clemex Imagine Analysis System were used for the microstructural observation and quantitative phase analysis. Also, the microstructural details, type of fracture and composition of inclusions were determined by Tescan Vega II XMU scanning electron microscopy linked to a Rontec EDS system.
Table 3 The electrical parameters of the welding process. Voltage (V)
Current (A)
Welding speed (cm/min)
Stick out (mm)
30
400
25.4
28
Fig. 1. SEM micrographs of the HAZ microstructure: (a) coarse grained zone; (b) fine grained zone.
B. Beidokhti et al. / Journal of Materials Processing Technology 209 (2009) 4027–4035
4029
Table 4 Quantitative microstructural analysis. Weld number
Acicular ferrite (AF)%
Grain boundary Widmannstätten ferrite (GBF)% ferrite (WF)%
Other phases (B and ferrite with M/A)%
T00 T10 T20 T30 T40 T50
70.1 74.6 71.7 61.0 54.2 48.1
15.8 13.6 13.5 14.7 13.0 15.3
5.9 6.5 8.9 19.7 27.9 32.8
8.2 5.3 5.9 4.6 4.9 3.8
Fig. 3. The variation of: (a) Impact energy; (b) shear fraction area in different titanium contents of the weld metals.
Fig. 2. Hardness values vs. titanium percentage of the weld metals.
13.7% pearlite. The application of single pass welding and same base material created a similar heat affected zone (HAZ) in all welded specimens. The microstructures of HAZ were a mixture of ferrite and bainite in the coarse grained zone and equiaxed ferrite, pearlite
Fig. 4. Microstructure of the weld metals with different titanium contents: (a) T00, a mixture of acicular and grain boundary ferrite; (b and c) T10 and T20, increasing of acicular ferrite in the microstructure; (d) T30, bainitic region in the microstructure.
4030
B. Beidokhti et al. / Journal of Materials Processing Technology 209 (2009) 4027–4035
Table 5 Tensile properties. Weld number
Yield strength (MPa)
Tensile strength (MPa)
Elongation (%)
T00 T10 T20 T30 T40 T50
570 597 612 647 768 795
659 681 758 809 933 947
21.2 22.7 20.4 19.8 19.0 17.7
Table 6 Comparison of the inclusion chemical composition in different weld metals. Weld number
Ti (%)
Mn (%)
Si (%)
Al (%)
T00 T10 T20 T30 T40 T50
6.0 21.2 25.9 41.7 58.4 62.8
32.7 24.7 19.2 12.2 14.3 17.1
14.2 8.1 7.2 6.3 6.0 4.6
10.2 16.4 15.6 20.9 13.7 12.4
and a few M/A microconstituents in the fine grained zone, respectively (Fig. 1). The change of each microstructure in the weld metals with different level of alloying elements is shown in Table 4. For all weld metals, evaluations were performed to determine the amounts of acicular ferrite (AF), grain-boundary ferrite (GBF), Widmanstätten ferrite (WF) and other phases like bainite (B) and M/A microconstituents within the weld metal. 3.2. Mechanical properties The hardness values were increased by addition of titanium to the weld metal as shown in Fig. 2. The yield strength, tensile strength and percentage of elongation data obtained in the longitudinal direction of the welds are summarized in Table 5. The longitudinal yield strength and tensile strength were in the range of 570–795 and 659–947 MPa, respectively. The variation of Charpy-V notch impact energy and the percentage of shear fraction area of the weld metals at two degrees, −10 and −30 ◦ C, are shown in Fig. 3. It is observed that the best impact properties were attained for 0.02–0.05% Ti in both cases. Fig. 5. SEM micrographs of the weld metals: (a) T00, acicular ferrite and grain boundary ferrite; (b) T10, rich acicular ferrite microstructure; (c) T50, formation of bainite.
3.3. Inclusion analysis Table 6 lists the variations of inclusion chemical composition for different weld metals. According to Bose-Filho et al. (2007) findings, increasing the titanium content of weld metal promoted the titanium levels of inclusions.
high in the specimens T40 and T50. Weld no. T30 with 0.09% Ti (and low notch toughness) showed the HIC susceptibility as same as the lower titanium and manganese content weld metals.
3.4. HIC test
4. Discussion
The values of CLR, CTR and CSR for each weld are presented in Table 7. According to the results, the cracking parameters were only
Comparison of the weld compositions showed that increased titanium content promoted the amounts of manganese and silicon in the weld metal. The basic-alumina flux has a basicity index of about 1.3. This flux produces about 0.5% Mn and 0.2% Si transfer to the weld deposit. With addition of Ti to the flux, the recovery of Mn and Si was increased and the amounts of these elements in the weld metal were promoted. In the welds with high contents of Ti, the difference of manganese percentage between the weld metal and base metal was about 0.7%. Therefore, in titanium additions greater than 0.09%, the positive effect of titanium is in contest with the negative effect due to the high manganese content of weld metal.
Table 7 The cracking parameters of HIC test for each weld metal. Weld number
CLR (%)
CTR (%)
CSR (%)
T00 T10 T20 T30 T40 T50
10.7 8.5 12.3 14.4 24.5 28.4
3.1 1.5 4.4 4.2 3.5 4.7
0.3 0.2 0.8 1.3 2.8 3.5
B. Beidokhti et al. / Journal of Materials Processing Technology 209 (2009) 4027–4035
4031
Fig. 6. Fracture mode of specimens at −10 ◦ C: (a) T00; (b) T10; (c) T30; (d) T40.
It can be seen from the corresponding micrographs shown in Fig. 4 that a mixed microstructure with acicular ferrite (AF), grain-boundary ferrite (GBF) and Widmanstätten ferrite (WF) was obtained in each specimen. In the welds with 0.09% Ti and more, bainite and ferrite with secondary phase (M/A) were also found in the microstructure. This phenomenon is related to the high amount of manganese recovery. Also according to Evans (1986, 1992) for the titanium content greater than 30 ppm, the weld metal contains M/A constituents as a microphase. In the specimens with titanium content up to 0.05%, formation of acicular ferrite was promoted. Therefore, due to a competitive growth of acicular ferrite and Widmanstätten ferrite, the amount of Widmanstätten ferrite was decreased. Spontaneously, M/A constituents were promoted because the amount of alloying elements and hardenability were increased. The results of hardness evaluation correlated to the microstructure observations. When titanium was added to the weld metal, hardness values were increased; but up to 0.05% Ti, these values were lower than the acceptance criterion (248 HV) for sour service in accordance with NACE MR0175. The titanium addition in the range of 0.02–0.05% in the weld metal gave the optimal hardness value and microstructure. Increased titanium content of the weld metal greater than 0.09%, leads to an increase in the amount of bainite and M/A microconstituents and promotes hardness values. In the specimen contained 0.22% Ti, the hardness value was about
321 HV. This hardness is larger than NACE restriction and therefore the risk of cracking in sour service is increased. Previous researchers (Bose-Filho et al., 2007) evaluated the hardness of bainite and low carbon martensite in the weld metal of HSLA pipeline steel to be approximately 325 HV and 340 HV, respectively. The weld metal contained 0.09% Ti exhibited the hardness value of about 286 HV; therefore, formation of martensite phase in this specimen was less pronounced. On the other hand, if the flux with basicity index of greater than 2 is used, the manganese content of weld metal will be reduced. Then, it is predicted that the optimal properties can be obtained in as-deposited welds with higher Ti content, e.g. at 0.09% Ti. Manganese is an important alloying element for solid solution strengthening; also, manganese can affect the transformation of austenite during cooling from high temperatures. It is observed that addition of titanium increases manganese values in the weld metals. Increased manganese content promoted formation of hard phases, especially bainite in the microstructure of weld metals with Ti ≥ 0.09%. Fig. 5 shows SEM micrographs of the different weld metal microstructures. The nucleation of acicular ferrite was increased by addition of titanium (Fig. 5a and b); beyond the optimal range of titanium (0.02–0.05%), increased manganese of the weld metal was a dominant factor to determining the weld metal microstructure. Therefore, bainite was observed in the large areas of the weld metal T50 (Fig. 5c).
4032
B. Beidokhti et al. / Journal of Materials Processing Technology 209 (2009) 4027–4035
Fig. 7. Fracture mode of specimens at −30 ◦ C: (a) T00; (b) T10; (c) T30; (d) T40. The quasi-cleavage fracture extended in the larger regions of all specimens in comparison to −10 ◦ C ones.
Also, increased titanium content leads to an increase in the yield and also in the tensile stresses. In the low titanium weld metals, the elongation percentage was promoted; but when B and M/A appeared in the microstructure, this parameter was dropped. Decreased elongation of the 0.05% Ti weld metal, with a good distribution of acicular ferrite, was only 0.8% in comparison to the 0.004% Ti weld metal. The results of tensile test exhibited that Ti addition had no strong negative effect on the elongation values; e.g. the weld metal with titanium content of about 0.19% (and consequently 2.23% Mn) exhibited the elongation value of about 19%. Generally, the strengthening and improvement of HIC resistance for this AF pipeline steel were obtained by addition of titanium. In submerged arc deposits, the fine oxide inclusion dispersion might act as sites for TiN and TiC precipitation. The precipitated titanium carbonitrides are expected to provide dispersion strengthening and to act as beneficial hydrogen traps in hydrogen or hydrogen sulfide environments. However, the yield strength and tensile strength of the weld metal were increased by further addition of alloying elements and also the elongation values were acceptable; but the negative effect due to the formation of B and M/A appeared in the Charpy impact test. The Charpy impact results confirmed previous findings. The addition of Ti up to 0.05% increased the impact energy value and the percentage of shear fracture area. Then, according to the forma-
tion of bainite and ferrite with M/A, these factors were decreased. At −10 ◦ C, the highest impact energy of about 137 J was obtained in the weld metal with 0.02% titanium. Also, the 0.05% Ti weld metal gave the impact properties superior to the reference specimen (weld no. T00). In the higher titanium content specimens, impact energy was decreased and reached to 28 J for the weld metal T50 with 0.22% Ti–2.29% Mn. Scanning electron microscopy analysis of the fractured impact specimens were undertaken to characterize the mode of fracture in the weld metals. Figs. 6 and 7 show the fractured surface micrographs with the wide ranges of fracture mode at −10 and −30 ◦ C, respectively. The evolution of the fracture mechanism was evident from dimpled ductile to quasi-cleavage mode with increased the amounts of titanium and manganese. This condition revealed the high degree of hydrogen embrittlement risk in sour environments (the absorbed energy of about 23 J for the 0.22% Ti specimen at −30 ◦ C). The impact energies and the shear fracture areas were impressively decreased at −30 ◦ C. In the weld metals with titanium contents of about 0.02 and 0.05%, the impact energy values were 116 and 101 J, respectively; whereas, this value for the reference specimen was only 78 J. These findings imply that the positive effect of Ti on the impact properties is increased with decreased temperature. Increasing the amount of acicular ferrite due to titanium addition could be responsible for this improvement. The formation of hard
B. Beidokhti et al. / Journal of Materials Processing Technology 209 (2009) 4027–4035
4033
phases like B and M/A due to increased manganese of the high alloy weld metals, leads to impairment of toughness again. The interesting result was the lack of any great difference between impact energies of the weld metals with titanium additions in the range of 0.09–0.22%. Increased titanium content in this range leads to progressive increasing of manganese values in the weld metals. It seems under this condition, manganese is the controlling parameter of impact properties. Because the amount of manganese in the specimens are high enough and the variations of manganese content in this range are negligible, these weld metals have almost the same impact toughness. Decreased impact energies in the high titanium content weld metals could be the result of two conditions: 1. Increasing the manganese content in the weld metal resulted from the recovery of manganese. 2. The strengthening effect of titanium due to the excessive distribution of fine Ti-base particles in the microstructure. For attaining different titanium contents of the weld metals without impressive change in the manganese content, welding with high basicity index flux is recommended. The non-metallic inclusions observed in this work are the result of deoxidation and/or desulphurization reactions taking place in the weld pool. It has been shown by some researchers (Abson et al., 1978; Cochrane and Kirkwood, 1978; Ricks et al., 1982) that these inclusions are very important in the microstructural development of ferritic weld metals since they act as nuclei for acicular ferrite. It has also been suggested that inclusions may pin austenite grain boundaries and reduce their grain size. These inclusions may also be responsible for determining the fracture resistance of HSLA steel weld metals (Bose-Filho, 1995). Garbarz et al. (2003) investigated on the type of sulfide inclusions in steels. Inclusions were identified by microanalysis in thin foils as MnS, (Mn, Cu)S, CuS and Cu2 S. Bose-Filho et al. (2007) showed that inclusions of the weld metal with a classical microstructure were mainly MnO or/and MnS and SiO2 for a low titanium content. Increasing the titanium content of the weld led to an increase in titanium within the inclusions, and at 230 ppm or more of titanium in the weld, Ti2 O3 and MnO and/or MnS were the main inclusion constituents. In the weld metals with very high titanium content, 700 ppm, the amount of titanium in the inclusions varied in the range of 60–70%. Also Yuan (2007) found that Ti-stabilized ultra low carbon steels contained a large number of fine Ti(C, N) and coarse Ti4 C2 S2 precipitates. According to Bose-Filho et al. (2007), the results of our paper showed an increase in the titanium content of inclusions with increased titanium content of the weld metal. This phenomenon leads to change in the nature of inclusions and decreases the amount of Mn-rich inclusions such as MnS. Bose-Filho et al. (2007) examined this phenomenon at the lower percentage of titanium addition and higher alloying elements. Also, they did not determine the optimum percentage of titanium. The Ti-base inclusions improve impact toughness by increasing the formation of acicular ferrite in the microstructure. North et al. (1979) showed that titanium and vanadium promote the acicular ferrite formation in the CMnCb deposits, but aluminum, zirconium and rare earth metals do not. The principal driving forces for growth of a ferrite crystal from austenite comprise the difference in chemical free energy between ferrite and austenite, and the tendency for the system to minimize the total interfacial free energy of interface boundaries (Aronson, 1960). In the case of TiC and TiN, close matching between the (1 0 0) planes of ferrite and these particles occurs and the mismatch parameter (ı) is very small. For ı < 0.05 the interface is coherent, and for 0.05 < ı < 0.25 a semi-coherent boundary
Fig. 8. (a) HIC crack in the mid-thickness of base metal; (b) SOHIC crack in the bainitic weld metal.
is produced (Chadwick, 1972). Since coherent interface has a much less interfacial energy than semi-coherent interface, therefore TiC and TiN are favorable sites for ferrite lath nucleation. According to these findings, the results of our work showed that the addition of titanium up to 0.05% increased the formation of acicular ferrite in the microstructure; but increased manganese amounts in the high titanium content weld metals hindered development of acicular ferrite and encouraged formation of bainite and ferrite with M/A in the microstructure (Table 4). Also, several authors (Garet et al., 1998; Lee and Lee, 1984, 1987) have shown that MnS and TiC inclusions are temporary and permanent traps for hydrogen atoms, respectively. Consequently, the substitution of manganese inclusion with titanium one promotes the number of permanent trap sites for hydrogen. It is predicted that this fact improves the resistance of steel in the sour environment. The HIC investigation was done to determine the crack susceptibility of different specimens in the hydrogen containing environments. The microscopic studies showed that the HIC cracks were propagated significantly in the base metal; only the specimens T40 and T50 exhibited cracking in the weld metal (Fig. 8). These specimens had a very high hardness and low Charpy toughness as result of the high percentages of titanium and manganese in the weld metal. No crack was observed in the weld metal T30 (0.09% Ti) contained a relatively high percentage of bainite and M/A microconstituents. As was mentioned before, if the flux with basicity index of greater than 2 is used, the manganese content of weld metal will be reduced. Then, it is predicted that the good impact toughness and HIC resistance can be obtained in the as-deposited welds with higher Ti content (the 0.09% Ti weld metal, coincidentally, showed low toughness and no HIC susceptibility). The CLR value less than 15% was adopted as an acceptance criterion of HIC test according to Kushida et al. (1997). On the basis of the CLR values, only the weld metals T40 and T50 are not acceptable.
4034
B. Beidokhti et al. / Journal of Materials Processing Technology 209 (2009) 4027–4035
All specimens exhibited the same CLR, CTR and CSR values; but, the weld metal cracking increased these values in the specimens T40 and T50. In the base metal, the HIC cracks were observed along the segregation band in the mid-thickness as shown in Fig. 8a. The morphology of the weld metal cracks was different. According to previous findings (Carneiro et al., 2003; Kane and Cayard, 1998), the short cracks normally associated with HIC can form a stacked array in the through thickness direction under the influence of residual tensile stress. This phenomenon is referred to as stress oriented hydrogen induced cracking (SOHIC). In the specimens T40 and T50, the SOHIC cracks were seen in the weld metals containing high amount of bainite (Fig. 8b). These results revealed no HIC susceptibility in the weld metals with titanium contents less than 0.09%; but, it is impossible to find which one has the highest HIC resistance. Furthermore, all specimens in current investigation have high yield strength (≥535 MPa); in conformity with NACE TM0177 for sour service application, sulfide stress cracking (SSC) studies will be requiring in the future studies.
in all weld metals, except in the weld metals with high titanium content as much as 0.19%. The SOHIC type cracks were formed in the bainitic microstructure of these welds. 7. MnS and TiC inclusions are temporary and permanent traps for hydrogen atoms, respectively. The substitution of manganese inclusion with titanium one, promotes the number of permanent trap sites for hydrogen and improves the resistance of steel in sour environments. Acknowledgements The authors are thankful to National Iranian Gas Company (NIGC), for financial supporting of this project and SAFA Co., Saveh, for providing the materials and laboratory facility in their factory. Thanks are also due to A. Erfanfar, R. Shamloo, A. Zandipoor and A. Khoshakhlagh of SAFA Co., for their assistance and critical review of the work and J. Woolington and N. Beidokhti for assistance with proofreading. References
5. Conclusions The effect of titanium addition on the microstructure and toughness of weld metals in an API 5L-X70 pipeline steel produced by submerged arc welding technique has been investigated. Based on the obtained results, the conclusions are as follows: 1. With addition of Ti to the flux, the recovery of Mn and Si were increased and the amounts of these elements in the weld metal were promoted. In the specimens with high titanium contents, the difference of manganese percentage between weld metal and base metal was about 0.7%. Therefore, in the titanium addition greater than 0.09%, the positive effect of Ti is in contest with the negative effect due to the high Mn content of the weld metal. 2. As a result of increasing the titanium content, the amount of acicular ferrite was increased in the weld metal. With further addition of titanium, the microstructure has changed from a mixture of acicular ferrite, grain-boundary ferrite and Widmanstätten ferrite to a mixture of acicular ferrite, grainboundary ferrite, bainite and ferrite with M/A microconstituents. The recovery of manganese and the hardening effect of this element in the high titanium content weld metals, encouraged formation of hard phases like B and M/A in the weld microstructure. 3. The addition of Ti up to 0.05% increased the impact energy value and the percentage of shear fracture area. Afterward, according to the formation of harder phases such as B and M/A, these factors were decreased. The evolution of the fracture mechanism was evident from dimpled ductile to quasi-cleavage mode with increasing the amount of titanium and manganese. The positive effect of Ti on the impact properties increased with decreased temperature. 4. The best combination of microstructure and impact properties was obtained in the range of 0.02–0.05% Ti. The hardness values of these specimens were lower than the acceptance criterion (248 HV) for sour services in accordance to NACE MR0175. 5. The results showed an increase in the titanium content of inclusions with increasing the titanium content of the weld metal. This phenomenon causes a change in the nature of inclusions and decreases the amount of Mn-rich inclusions such as MnS. The Ti-base inclusions improve impact toughness by increasing the formation of acicular ferrite in the microstructure. 6. In the base metal, the HIC cracks were observed along the segregation band in the mid-thickness. No HIC crack was observed
Abson, D.J., Dolby, R.E., Hart, P.H.M., 1978. The role of nonmetallic inclusions in ferrite nucleation in carbon steel weld metals. In: International Conference on Trends in Steels and Consumables for Welding, London, UK, pp. 75–89. Aronson, H.J., 1960. The proeutectoid ferrite and proeutectoid cementite reactions. In: Met. Soc. AIME Symposium on Decomposition of Austenite by Diffusional Processes, Philadelphia, pp. 387–546. Bhole, S.D., Nemade, J.B., Collins, L., Liu, C., 2006. Effect of nickel and molybdenum additions on weld metal toughness in a submerged arc welded HSLA line-pipe steel. J. Mater. Process. Technol. 173, 92–100. Bose-Filho, W.W., 1995. Micromechanisms of brittle fracture in high strength low alloy steel weld metals. Ph.D. Thesis. The University of Birmingham, Birmingham, UK. Bose-Filho, W.W., Carvalho, A.L.M., Strangwood, M., 2007. Effect of alloying elements on the microstructure and inclusion formation in HSLA multipass welds. Mater. Charact. 58 (1), 29–39. Carneiro, R.A., Ratnapuli, R.C., Lins, V.F.C., 2003. The influence of chemical composition and microstructure of API linepipe steels on hydrogen induced cracking and sulfide stress corrosion cracking. Mater. Sci. Eng. A 357, 104–110. Chadwick, G.A., 1972. Metallography of Phase Transformations. Butterworth Publications, London, pp. 193–200. Cochrane, R.C., Kirkwood, P.R., 1978. The effects of oxygen on weld metal microstructure. In: The Welding Institute Conference on Trends in Steels and Consumables, London, UK, pp. 103–122. Contreras, A., Albiter, A., Salazar, M., Perez, R., 2005. Slow strain rate corrosion and fracture characteristics of X-52 and X-70 pipeline steels. Mater. Sci. Eng. A 407, 45–52. Crockett, D.D., Rhone, J.A., Young, R.F., Noernberg, D.C., 1995. Design considerations for submerged arc consumables intended for the manufacture of line pipe. Pipeline Technol. 1, 151–162. Evans, G.M., 1986. The effect of silicon on the microstructure and properties of C–Mn all-weld metal deposits. Met. Constr. 28, 438R–444R. Evans, G.M., 1992. Effect of micro-alloying elements in C–Mn steel weld metal. In: ADESOL Conference, Madrid, Spain. Evans, G.M., 1998. Effect of molybdenum on microstructure and properties of C–Mn all-weld metal deposits. Join. Mater. 1 (5), 239–246. Garbarz, B., Marcisz, J., Wojtas, J., 2003. TEM analysis of fine sulphides dissolution and precipitation in steel. Mater. Chem. Phys. 81, 486–489. Garet, M., Brass, A.M., Haut, C., Guttierez-Solana, F., 1998. Hydrogen trapping on non metallic inclusions in Cr–Mo low alloy steels. Corros. Sci. 40, 1073–1086. Gladman, T., 1997. The development of microalloyed steels. In: The Physical Metallurgy of Microalloyed Steels, The Institute of Materials, London, England, pp. 1–18. Jorge, J.C.F., Souza, L.F.G., Rebello, J.M.A., 2001. The effect of chromium on the microstructure/toughness relationship of C–Mn weld metal deposits. Mater. Charact. 47, 195–205. Junhua, K., Lin, Z., Bin, G., Pinghe, L., Aihua, W., Changsheng, X., 2004. Influence of Mo content on microstructure and mechanical properties of high strength pipeline steel. Mater. Des. 25, 723–728. Kane, R.D., Cayard, M.S., 1998. Roles of H2 S in the behavior of engineering alloys: a review of literature and experience. In: NACE, Corrosion/98. Paper No. 274. Houston, USA. Keehan, E., Andren, H.O., Karlsson, L., Murugananth, M., Bhadeshia, H.K.D.H., 2002. Microstructural and mechanical effects of nickel and manganese on high strength steel weld metals. In: Sixth International Conference: Trend in Welding Research, Pine Mountain, GA, USA, pp. 695–700. Kim, S.H., Bang, C.Y., Bang, K.S., 2001. Weld metal impact toughness of electron beam welded 9% Ni steel. J. Mater. Sci. 36, 1197–2000.
B. Beidokhti et al. / Journal of Materials Processing Technology 209 (2009) 4027–4035 Kushida, T., Okaguchi, S., Hamada, M., Yamamoto, A., Ohnishi, K., Fujino, J., 1997. Study of X80 grade high strength line pipe for sour service. In: NACE, Corrosion/97. Paper No. 24. Houston, USA. Lee, H.G., Lee, J.Y., 1984. Hydrogen trapping by TiC particles in iron. Acta Metall. 32, 131–136. Lee, J.Y., Lee, J.L., 1987. A trapping theory of hydrogen in pure iron. Philos. Mag. A 56, 293–309. Lee, W.B., Hong, S.G., Park, C.G., Kim, K.H., Park, S.H., 2000. Influence of Mo on precipitation hardening in hot rolled HSLA steels containing Nb. Scripta Mater. 43, 319–324. North, T.H., Bell, H.B., Koukabi, A., Craig, I., 1979. Notch toughness of low oxygen content submerged arc deposits. Weld. J. 58 (12), 343s–354s. Pickering, F.B., 1983. The spectrum of microalloyed high strength low alloy steels. In: HSLA Steels Technology and Applications, Conference Proceedings of International Conference on Technology and Applications of HSLA Steels, pp. 1–32. Ricks, R.A., Barrite, G.S., Howell, P.R., 1982. The nature of acicular ferrite in HSLA steel weld metals. J. Mater. Sci. 17, 732–740.
4035
Sun, W., Lu, C., Tieu, A.K., Jiang, Z., Liu, X., Wang, G., 2002. Influence of Nb, V and Ti on peak strain of deformed austenite in Mo-based micro-alloyed steels. Mater. Process. Technol. 125/126, 72–76. Yuan, X., 2007. Precipitates and hydrogen permeation behavior in ultra-low carbon steel. Mater. Sci. Eng. A 452/453, 116–120. Zhao, M.C., Yang, K., 2005. Strengthening and improvement of sulfide stress cracking resistance in acicular ferrite pipeline steels by nano-sized carbonitrides. Scripta Mater. 52, 881–886. Zhao, M.C., Yang, K., Shan, Y., 2002. The effects of thermo-mechanical control process on microstructures and mechanical properties of a commercial pipeline steel. Mater. Sci. Eng. A 335, 14–20. Zhao, M.C., Yang, K., Shan, Y., 2003. Continuous cooling transformation of undeformed and deformed low carbon pipeline steels. Mater. Sci. Eng. A 335, 126– 136. Zhong, Y., Xiao, F., Zhang, J., Shan, Y., Wang, W., Yang, K., 2006. In situ TEM study of the effect of M/A films at grain boundaries on crack propagation in an ultra-fine acicular ferrite pipeline steel. Acta Mater. 54, 435–443.