Proceedings of the Eighteenth (2008) International Offshore and Polar Engineering Conference Vancouver, BC, Canada, July 6-11,

2008 Copyright 2008 by The International Society of Offshore and Polar Engineers (ISOPE) ISBN 978-1-880653-70-8 (Set); ISBN 1-880653-68-0 (Set)

DWTT properties for high strength line pipe steels

Takuya HARA, Yasuhiro SHINOHARA, Yoshio TERADA and Hitoshi ASAHI
Technical Development Bureau, Nippon Steel Corporation Kimitsu city, Chiba, Japan

ABSTRACT
High strength line pipe steels needs to have good crack arrestability to arrest cracks in the pipe body even if brittle fracture occurs at welds, or running ductile fractures occur in the pipe body. The drop weight tear test (DWTT) is one of the major test methods to evaluate crack arrestability of brittle and running ductile fractures. In order to obtain a good DWTT shear area for high strength line pipe steels, it is important to clarify the conditions of brittle fracture occurrence because the shear area decreased remarkably once brittle fracture occurs. In this study, the conditions of brittle fracture occurrence such as slant fracture for high strength line pipe steels were investigated. Separation was formed when a ferrite and bainite dual-phase microstructure was formed. Slant fracture was formed when a bainite single phase was formed. The suppression of slant fracture and remarkable separation is important for improving DWTT properties. Excellent DWTT properties were obtained for X100 line pipe steel with respective thicknesses of 19 mm, 16 mm and 14 mm.

area of more than 40%. However, a DWTT shear area of 85% or higher is required for specifications such as those of the API (American Petroleum Institute) because DWTT shear area scattering taken into account in a circumferential direction. The key problem when evaluating crack arrestability of brittle fracture using DWTT is the fact that brittle fracture does not often occur just under the notch. Because high strength line pipe steels involve low carbon chemistry and clean and pure steels, it is very difficult to initiate brittle fracture under the notch, thus invalidating this test. To solve this problem, the precrack DWTT or Chevron notch DWTT are proposed (Wilkowski et al. (1978) and Maxey et al. (1985)). For the former, it is easy to initiate brittle fracture after the ductile fracture is initiated using a three point bend test, whereas for the Chevron notch DWTT, because the notch dimension differs from the press notch DWTT, a brittle fracture is easy to initiate due to severe stress constraint. Conversely, the required Charpy V-notch energy (Maxey et al. (1976), Subcommitee Summary report-AISI (1974) and Poynton (1974)) for crack arrestability of running ductile fracture was proposed in comparison with the full crack burst test results and the Charpy Vnotch energy of the line pipe body, and the Charpy V-notch energy correlated with that of the full crack burst test results up to X70 grade. However, the safety factor such as 1.3 or 1.7 was multiplied to the required Charpy V-notch energy for high strength steels with a grade of X80 or higher because the Charpy V-notch energy did not correlate with the full crack burst test results (Demofonti et al. (2004)). Likewise, for the X100 or X120 grades, the required energy multiplied by the safety factor did not correlate with the full crack burst test results. Based on these results, the HLP (High grade Line Pipe) Committee for the ISIJ (Iron and Steel Institute of Japan) proposed that precrack DWTT energy was used as the required energy for crack arrest of a running ductile fracture instead of Charpy V-notch energy (Sugie (1988) and Makino (2001)). CSM (Centro Sviluppo Materiali) also proposed that crack propagation energy defined as the integral of load vs. displacement after the maximum load for the DWTT was suitable for the required energy for crack arrest of a running ductile fracture (Demofonti et al. (2004)). Another research group proposed that the CTOA (Crack Tip Opening Angle) is effective for evaluating a running ductile fracture (Fonzo et al. (2002)). Therefore, the DWTT is the important method used to evaluate the

KEYWORDS: High strength line pipe, DWTT, separation, slant

fracture, texture, ferrite, bainite

INTRODUCTION
Recently, high strength line pipe steels, with yield strength of X80 or higher, have been used for many pipeline projects because of the reduced cost for the transportation of natural gas. Crack arrestability of brittle and running ductile fractures is needed as one of the required properties for high strength line pipe steels. For example, cracks must be arrested, even if brittle fracture occurs from the welds such as girth welds. Cracks must also be arrested if the line pipe body is subject to ductile fracture. The DWTT (Drop Weight Tear Test) (Eiber et al. (1979)) is one of the major test methods used to evaluate the crack arrestability of brittle fracture. This test evaluates whether a ductile crack is transferred from a brittle fracture after a brittle crack is initiated just under the notch. Previous results (Amano et al. (1986)) indicated that the crack speed became lower than 450 m/s and that the crack was subsequently arrested in the full crack burst test, for line pipes with a DWTT shear

189

crack arrestability of brittle and running ductile fractures. Improving the DWTT energy and DWTT shear area is important in obtaining good crack arrestability of brittle and running ductile fractures. Namely, it is important to grasp the brittle fracture morphology in order to suppress brittle fracture. In this paper, the effects of microstructure and texture on DWTT brittle fracture were investigated for X100 high strength line pipe steels in order to clarify the condition of brittle fracture occurrence for high strength line pipe steels. The metallurgical concept to improve DWTT properties is also described.

Tensile strength increased with decreasing the start cooling temperature as indicated in Figs. 1 and 2. The reason that tensile strength increased as the start cooling temperature decreased is thought to be dependent on the stop cooling temperature. The morphology of the second phase except the ferrite phase was remarkably influenced by the stop cooling temperature. Namely, tensile strength tends to increase with decreasing the stop cooling temperature.
100

DWTT shear area, /%

80 60 40 20 X100 0 500 600 700 800

EXPERIMENTAL PROCEDURE
Table 1 shows the chemical compositions of the tested steel. A 300 ton continuous casting slab of X100 line pipe steel with a carbon content of 0.06% was melted in a converter and continuous cast. The continuous casting slab of width 280 mm, length 250 mm, and thickness 240 mm was rolled in a laboratory. Table 2 shows the TMCP conditions. The slab, reheated at 1150 C, was rolled in 23mm thick plate and then cooled and the start cooling temperature was changed from 736 to 540 C. Start and stop cooling temperature at surface was measured using an emission thermometer. A full thickness tensile specimen and DWTT specimen were taken from the plate in a transverse direction. The tensile specimen was a JIS No. 5 full thickness specimen, and the DWTT specimen was prepared in accordance with API 5L/ISO3183 specifications. The DWTT was carried out at -20 C and the DWTT energy and DWTT shear area were measured. The microstructure of the X100 line pipe steel was observed using SEM (Scanning Electron Microscopy) and EBSP (Electron Back Scattering Pattern). The {100} texture was also measured using X-ray analysis and EBSP. Table 1 Chemical composition of tested steels. (mass %)

Start cooling temp., /oC

Fig. 1 Effect of start cooling temperature on DWTT shear area.

100

DWTT shear area, /%

80 60 40 20

X100 TS
0 700 750 800 850

X100 900

C X100 0.06

Mn 1.90

Mo Ti Others Remarks 0.26 0.015 Ni, Cu, Nb CC slab

Table 2 TMCP conditions of laboratory rolling.

Steel IDQ1 IDQ2 DLQ1 DLQ2 DLQ3 DLQ4 Finish rolling Reheating Start temp Stop temp Temp. Start temp Finish temp Water cooling Water cooling o o o o o C C C C C 736 300/400 736 450/500 660 200/300 1150 850 800 638 RT 585 RT 540 RT

Tensile strength, /MPa Fig. 2 Relationship between DWTT shear area and tensile strength.

Effect of the start cooling temperature on brittle fracture morphology Figure 3 shows the effect of the start cooling temperature on brittle fracture morphology. A slant fracture was formed at a start cooling temperature of 730 C, while conversely, remarkable separation was formed at start cooling temperatures of 600 and 550 C and both slant fracture and separation were formed at a start cooling temperature of 650 C. As for the DWTT fracture appearance for X100 line pipe steel, slant fracture and separation were observed. Separation was defined as a fracture morphology where many cracks were formed parallel to the rolling plane and slant fracture was defined as a fracture morphology, where brittle fracture was formed along the plane rotated 40 from the rolling plane on the axis of the rolling direction from two plane analysis. Effect of the start cooling temperature on the microstructure Figure 4 shows the effect of the start cooling temperature on the microstructure. A bainite single phase was formed at a start cooling temperature of 730 C, while conversely, a ferrite and bainite dual phase microstructure was formed below a start cooling temperature of 650 C. Figure 5 shows the effect of the start cooling temperature on the ferrite fraction. The ferrite fraction increased as the start cooling temperature went down. Effect of the start cooling temperature on the texture

EXPERIMENTAL RESULTS
Effect of the start cooling temperature on DWTT properties Figure 1 shows the effect of the start cooling temperature on DWTT shear area. The DWTT shear area was about 60 to 70% with a start cooling temperature of 730 C. The DWTT shear area increased with decreasing the start cooling temperature, and a DWTT shear area above 80% was obtained with a start cooling temperature below 650C. Figure 2 shows the relationship between DWTT shear area and tensile strength. All data were satisfied with tensile strength of X100 grade in this study. The DWTT shear area increased regardless of increasing tensile strength as indicated in Fig. 2 although the DWTT shear area normally decreased with increasing tensile strength.

190

Ferrite fraction, /%

Figure 6 shows the effect of the start cooling temperature on the texture. An intensity of {100} of the plane parallel to the rolling plane increased with decreasing the start cooling temperature. Conversely, an intensity of {100} of the plane rotated 40 from the rolling plane on the axis of the rolling direction did not remarkably change irrespective of the start cooling temperature as compared with an intensity of {100} of the plane parallel to the rolling plane. Strictly speaking, an intensity of {100} of the plane rotated 40 from the rolling plane very slightly increased with increasing the start cooling temperature.
IDQ1 (736oC) DLQ2 (638oC) DLQ4 (540oC)

50 40 30 20 10 0 500 600 700

o

X100

800

Start cooling temp., / C Fig. 5 Effect of start cooling temperature on the ferrite fraction.
{100} intensity vs random
3

Fig. 3 Effect of start cooling temperature on DWTT fracture appearance at 20oC.

IDQ1 (736oC)

1 500

plane parallel to RD plane rotated 40o from RD 600 700

o

800

Start cooling temp., / C Fig. 6 Effect of start cooling temperature on {100} intensity. (RD : rolling direction)

DISCUSSION
Condition of separation occurrence for high strength line pipe steels The condition of separation occurrence for high strength line pipe steels was summerized as a comparison with a previous report concerning the condition of separation occurrence (Matsuda et al (1982)). Separation was formed at a start cooling temperature below 650 C, and was formed for X100 line pipe steel with a ferrite and bainite dual phase microstructure. On the other hand, no separation occurred at a start cooling temperature of 730 C and no ferrite was formed at a start cooling temperature of 730 C, so separation was thought to be formed when the ferrite and bainite dual phase microstructure was formed. {100}, which corresponds to cleavage plane in iron, was integrated to the plane parallel to the rolling plane as the start cooling temperature decreased although high intensity of {100} was observed in the plane rotated 40o from the rolling plane on the axis of the rolling direction, as indicated in Fig. 6. There are many interfaces between a soft phase such as ferrite and a hard phase such as M-A (Martensite-Austenite Constituent) in dual phase microstructure. Brittle fracture becomes easily occurs from the interface between the soft and hard phases, and subsequently propagates along the interface between the soft and hard phases or along {100} parallel to the rolling plane. Figure 7 shows the relationship between crack propagation for separation and the microstructure of DLQ4 using an EBSP (Electron back scattering pattern). The cross section of transverse direction under the DWTT press notch was polished and EBSP measurement was carried out. There are many {100} in the cracks propagated along the interface, as indicated by some circles in Fig. 7. This {100} coincided with the ferrite phase as indicated by some circles in Fig. 7. This result corresponds to a high integrity of {100} to a plane parallel to the rolling plane for a dual phase microstructure as indicated in Fig. 6. Therefore, it is thought that separation was formed when brittle fracture occurred

DLQ2 (638oC)

DLQ4 (540oC)

20m

Fig. 4 Effect of start cooling temperature on microstructure.

191

at the interface between the soft and hard phases and propagated along the interface or along {100} parallel to the rolling plane for the dual phase microstructure.

The crack propagated along the plane rotated 40 from the rolling plane after brittle fracture occurred from the brittle phase such as M-A or segregation. Figure 9 shows orientation image mapping on the plane rotated 40 from the rolling plane and many {100} was integrated to this plane, which indicates that a crack can easily propagate on this plane. A slant fracture was formed by the {100} integration to the plane rotated 40 from the rolling plane due to the bainite transformation texture from deformed austenite (Inagaki et al. (1973)).

Fig. 7 Relationship between crack propagation for separation and microstructure of DLQ4. (Red mark shows {100}. ) Matsuda et al. (1982) reported details of the condition of separation occurrence using X70 line pipe steels. This report indicated the condition of separation occurrence, mainly concerning deformed ferrite microstructure. In this case, {100} and {111} were integrated in the deformed ferrite and brittle fracture occurring from the interface between {100} and {111}, while the crack propagated along {100}. Matsudas results differ from these results in the point of microstructure. In other words, Matsudas report indicated the condition of separation occurrence with reference to deformed ferrite, while conversely, this paper describes the condition of separation occurrence concerning the ferrite and bainite dual phase microstructure. Condition of slant fracture occurrence for high strength line pipe steels Next, the condition of slant fracture occurrence was considered for high strength line pipe steels. Slant fracture occurred at a start cooling temperature exceeding 660 C. A bainite single phase was formed in the case that the slant fracture was formed. Figure 8 shows a {200} pole figure of X100 line pipe steel cooled from a start cooling temperature of 730 C. These arrows indicated in Fig.8 show the plane corresponding to the slant fracture. The {100} was remarkably integrated to the plane rotated 40 from the rolling plane on the axis of the rolling direction however, {100} was not greatly integrated to the plane parallel to the rolling plane. The plane of the slant fracture coincided with the plane rotated 40 from the rolling plane on the axis of the rolling direction from two plane analysis.

Fig. 9 EBSP mapping showing the plane parallel to the slant fracture. Condition of brittle fracture occurrence for high strength line pipe steels A typical brittle fracture for X100 high strength line pipe steel was separation and slant fracture. The condition of brittle fracture for high strength line pipe steels is summarized in Fig. 10. Separation was formed when a dual phase microstructure was formed remarkably if a laminar dual phase microstructure was formed. Brittle fracture is easy to initiate and propagate readily because a significant interface of soft and hard phases is easily formed. Slant fracture was formed when a bainite single phase was formed and many {100} were integrated to the plane rotated 40 from the rolling plane. Here, it is thought that no slant fracture occurs when there is no brittle phase, even if a bainite single phase is formed, and {100} was integrated on the plane rotated 40 from the rolling plane.
High intensity of (100) rotated 40 from rolling plane

RD

Ferrite and bainite single phase*)

High intensity of (100) parallel to rolling plane

Bainite single phase

Low intensity of (100) parallel to rolling plane

TD

Separation*
*)MAs was formed along ferrite boundary

Slant fracture

Fig. 8 {200} pole figure of X100 line pipe steel cooled from a start cooling temperature of 730 C.

Fig. 10 Condition of brittle fracture occurrence for high strength line pipe steels.

192

Metallurgical design for the improvement of DWTT properties Finally, the metallurgical design for the improvement of DWTT properties is explained. First, the suppression of brittle fracture occurrence is important in order to improve the DWTT shear area. The DWTT shear area was drastically reduced by the formation of slant fracture. On the other hand, the DWTT shear area maintained a high value of 80% or higher, even if slight separation occurred. Namely, the formation of a microstructure without the formation of slant fracture is very effective for improving the DWTT shear area. The formation of a microstructure without remarkable laminar dual phase is also effective for DWTT properties, especially, DWTT propagation energy (HARA et al. (2004)). Using this metallurgical concept, excellent low temperature toughness of X100 line pipe steels with thicknesses of 14 mm, 16 mm and 19 mm respectively has been developed. For example, Fig. 11 shows shear area different types of DWTT for X100 line pipe steel with 16 mm thick. A high shear area of 85% or higher was obtained down to 50C irrespective of different notch DWTT.
100

REFERENCE
Amano, K, Matsuoka, M, Ishihara, T, Tanaka, K, Inoue, T, Kawaguchi, Y, and Tsukamoto, M (1986). Significance of Yield Ratio Limitation to Plastic Deformation of Pipeline in High Pressure Proof Test, proceedings of the 7th Symposium on line pipe research, AGA, Houston, USA, pp 8-1. Demofonti, G, Mannucci, G, Biago, MDI, Hillenbrand, HG, and Harris, D (2004). Fracture Propagation Resistance Evaluation of X100 TMCP Steel Pipes for High-Pressure Gas Transportation Pipelines Using Full-scale Burst Tests, 4th International Conference on Pipeline Technology Conference, Ostend, Belgium, pp 467-476. Demofonti, G, Hillenbrand, HG, Mannucci, G, and Harris, G (2004). Evaluation of the Suitability of X100 Steel Pipes for High Pressure Gas Transportation Pipelines by Full Scale Test, proc. of Biennial IPC, Calgary, Alberta, Canada, pp 1685-1692. Eiber, RJ, and Maxey, WA (1979). Fracture Propagation Control Methods, proceedings, Annual Symposium, Society of Flight Test Engineers, pp 1-16. Fonzo, A, Salvini, P, Di, Biago, M, and Manuucci, G (2002). Full History Burst Test through Finite Element Analysis, proc. of IPC, Calgary, Alberta, Canada. Hara, T, Shinohara, Y, Asahi, H, and Terada, Y (2006). Effects of Microstructure and Texture on DWTT Properties for High Strength Line Pipe Steels, proc. of IPC, Calgary, Alberta, Canada. Inagaki, H, Kurihara, K, and Kozasu, I (1973). Influence of Crystallographic Texture on the Strength and Toughness of the Controlled Rolled High Tensile Strength Steel, Tetsu-to-Hagane, 61, 7, pp 991-1011. Makino, H, Kubo, T, Shiwaku, T, Endo, S, Inoue, T, Kawaguchi, Y, Matsumoto, Y, and Machida, S (2001). Prediction for Crack Propagation and Arrest of Shear Fracture in Ultra-High Pressure Natural Gas Pipelines, ISIJ International, 41, 4, pp 381-388. Matsuda, S, Kawashima, Y, Sekiguchi, S, and Okamoto, M (1982). Mechanism of Separation of a Low-Carbon Low Alloy Steel Control-Rolled in Austenite and Ferrite Two Phase Regions, Tetsuto-Hagane, 68, 3, pp 435-443. Maxey, WA, Koch FO, Peeck A, Junker G, Vogt GH, Peters PA, Kuegler, J, and Seifert, K (1985). Transition Temperature Tests Show Fracture Behavior, Oil and Gas Journal, 83, 16, pp 75-80. Maxey, WA, Kiefner, JF, and Eiber, RJ (1976). Final Report on Ductile Fracture Arrest in Gas Pipelines to Pipeline Research Committee American Gas Association, AGA Catalog No.L32176. Poynton, WA, Shannon, RWE, and Fearnehough, GD (1974). The Design and Application of Shear Fracture Propagation Studies, Trans ASME Ser H, 96, 4, pp 323-329. Subcommittee Summary Report-AISI committee of large diameter line pipe producers (1974). Running Shear Fracture in Line Pipes. Sugie, E, Matsuoka, M, Akiyama, T, Tanaka, K, and Tsukamoto, M (1988). Propagating Shear Fracture in Natural Gas Transmission Pipelines, ASTM STP945, pp 237-246. Wilkowski, GM, Maxey, WA, and Eiber, RJ (1978). Problems in Using the Charpy, Dynamic Tear Test and Drop Weight Tear Test for High Toughness Quenched and Tempered Steels, AIME annual meeting, Denver, Colorado, USA, pp 201-225.

DWTT Shear area, /%

80 60 40
Press notch DWTT

20
X100 line pipe 16mmWT

Precrack DWTT Chevron notch DWTT

0 -60 -50 -40 -30 -20

o

-10

Temperature,/ C Fig. 11 DWTT shear area transition curve for 16 mm thick X100 line pipe steels.

CONCLUSIONS
The effects of microstructure and texture on DWTT brittle fracture morphology were evaluated in order to clarify the conditions of brittle fracture occurrence for high strength line pipe steels of X80 or higher. The main conclusions were as follows: (1) Typical brittle fracture for the DWTT was separation and slant fracture for X100 line pipe steels. (2) Separation was formed when a ferrite and bainite dual phase microstructure was formed. Brittle fracture occurred from the interface between ferrite and the hard phase such as M-A, where the crack propagated along the interface between the soft and hard phases, which is parallel to the rolling plane or along {100} parallel to the rolling plane. (3) Slant fracture was formed when a bainite single phase was formed. Slant fracture occurred by integration of {100} to the plane rotated 40 from the rolling plane on the axis of the rolling direction after brittle fracture occurred from the brittle phase such as M-A or segregation. (4) The suppression of slant fracture and remarkable separation is important for improving DWTT properties. Excellent DWTT properties were obtained for X100 line pipe steel with thicknesses of 14mm, 16mm and 19mm.

193