(Concept of Development and Properties of 10, 25 and 40 mm Thick 7%Ni Steel Plate)

Transcription

1 [ 28 p (2010)] (Concept of Development and Properties of 10, 25 and 40 mm Thick 7%Ni Steel Plate) by KUBO Naoshige, TAKATA Masanori, YAMASHITA Maki, KNOWLES David, HIROSE Hitoshi, SAKATO Naokazu, MURAMOTO Seiichi, HIRAI Syuuichi, MITSUMOTO Masahiko, ARIMOCHI Kazushige, KAWABATA Tomoya and KAMO Takahiro LNG storage tanks made of 9%Ni steel plate have safely been operated at the many LNG export and import terminals in the world for the last half-century. Over the years extensive improvement and enhancement have been achieved on the 9%Ni steel plate in terms of toughness in steel to withstand the cryogenic temperatures under an operating condition with LNG. This paper reports research and development of the 7%Ni - TMCP steel plate. As nickel is an expensive and valuable rare metal, it was aimed to reduce nickel content to 7% to save investment costs for constructions of LNG storage tanks. Inferiority of the lower nickel content was compensated by adjusting chemical contents of Mn, Si, Cr etc. and by Thermo-Mechanical Controlled Process (TMCP) with an intermediate heat treatment. In order to evaluate fitness of the 7%Ni - TMCP steel plate and its weld for LNG storage tanks a series of testing was conducted. Three different kinds of plate thicknesses of 40, 25 and 10 mm were chosen to run large scale fracture toughness tests such as duplex ESSO tests, cruciform wide plate tests as well as small scale tests. Those mechanical tests showed excellent quality of the steel and demonstrated safety of the steel which was considered the same safety level as the conventional 9%Ni steel plate. Hence, it is considered that the 7%Ni - TMCP steel plate warrants serious consideration for use in LNG storage tanks. Key Words: LNG, tank, fracture, brittle, CTOD, 9%Ni steel, 7%Ni steel 1. Introduction As the global Liquefied Natural Gas (LNG) demand remains high, construction of LNG storage tanks is expected to continue to increase in the future too. 9%Ni steel plates 1 5) have been abundantly used for the primary component of aboveground LNG storage tanks for the last half-century. It is well known that an excellent cryogenic fracture toughness of the nickel alloyed steel plate is attributed to the retained austenite and refined microstructure obtained by nickel content and heat treatment process. The conventional 9%Ni steel plate 6) was developed to target the above characteristics. The newly developed 7%Ni- TMCP steel plate employs a Thermo-Mechanical Control Process (TMCP) to obtain the retained austenite and refined microstructure to reinforce the inferiority of two percent nickel content (Fig. 1). Reduction of nickel content significantly contributes to save natural resources as well as to mitigate the anticipated construction cost rising due to significant fluctuations in the nickel price. This paper describes basic concept of the development and mechanical properties of the 7%Ni-TMCP steel plate including welds. The paper also provides large-scale fracture toughness test results of the 7%Ni-TMCP steel plate to assess safety performance of the LNG storage tank comparing with the previous study results 7 10) of the 9%Ni steel plate constructed in the large capacity LNG storage tanks. * ** Osaka Gas Co., Ltd *** Osaka Gas Engineering Co., Ltd. **** Shell Global Solutions International B.V. ***** G H GH Engineering Consultants ****** Toyo Kanetsu K.K. ******* Member, Sumitomo Metal Industries, Ltd. ******** Sumitomo Metal Industries, Ltd. Fig. 1 Target of development.

2 Fig. 2 Requisite properties for material of LNG tank 11). Fig. 3 Improvement of CTOD property in welded joint by decrease of Si and Cr bearing. Fig. 4 Auto-temper phenomenon in HAZ by decrease of Si and Cr bearing. 2. Metallurgical Basis of 7%Ni-TMCP steel plate Resistance to brittle crack initiation especially in heat affected zone (HAZ) and arrest of propagating cracks in base plate were targeted as key properties to ensure the safe operation of LNG storage tanks (Fig. 2). Researches for effective chemical composition and production process found that superior low silicon and proper chromium alloying with grain refinement by Thermo-Mechanical Control Process leads to 2% nickel content saving. 2.1 Design of Chemical Compositions for Enhancement of Brittle Crack Initiation Resistance at the Toe of Welded Joint An early investigation with the Ni simply decreased from 9% to 7% showed significant deterioration in crack initiation property evaluated by three point bend CTOD (Crack Tip Opening Displacement) testing compliant with BS ) at the weld toe where GTAW method was employed with use of the same kind of welding consumable (70% nickel) as that used in the 9%Ni steel plate (Fig. 3). This was thought to be because i) microstructure just under toe is coarse grain HAZ without reheating by following passes, ii) sufficient hardenability cannot be obtained and iii) plastic deformation of weld metal composed of austenite can scarcely be expected unlike in the case of fusion line. To improve toughness of HAZ at the toe position in the 7%Ni steel welds, it was considered to be effective to adjust hardenability of the steel and tempering the microstructure of HAZ by following welding passes 13). "Tempering" is known as an effective process to precipitate cementite particles in the microstructure and mitigate the excess hardness of matrix to suitable level. As mentioned above microstructure just under toe is unable to be tempered in the absence of following passes. Therefore, chemical composition is to be designed so that microstructure of HAZ should be automatically tempered without following welding passes. Lowering Si content is effective to bring an auto-

3 132 Kubo etc. Development of 7%Ni-TMCP Steel Plate for LNG Storage tanks Table 1 Chemical compositions of steel used [mass%]. Fig. 5 Generals of production processes and typical microstructure. tempering effect on the matrix at the toe position (Fig. 4). Furthermore, adjustment of other chemical compositions especially Cr, which has a role to raise hardnability without inhibition of formation of retained austenite, is also effective. As a result of these studies, it was recognized that resistance to brittle crack initiation at the toe position of the developed steel with 7%Ni would be significantly improved and reach to the equivalent level to that of 9%Ni steel (Fig. 3). Finally, chemical compositions of developed steel have been obtained as shown in Table Application of TMCP for Enhancement of Arrestability of Propagating Brittle Crack Thermo-mechanical Controlled Process (TMCP) with large reduction just above Ar3 is known as an effective process for enhancement of toughness due to refinement of microstructure 6, 14). The TMCP with an addition of intermediate heat treatment could achieve finer microstructure for the 7%Ni steel production (Fig. 5). Retained austenite is also known to contribute for improving toughness of high nickel steels. As nickel depresses Ms transformation temperature and serves to stabilize austenite thermally, the decrease of nickel generally reduces retained austenite fraction. However, it is clearly shown that the 7%Ni steel plates manufactured by the TMCP with an intermediate heat treatment achieve higher retained austenite volume than that of Fig. 6 Evaluation result of retained austenite. 9%Ni steel (Fig. 6). 3. Mechanical Properties of 7%Ni-TMCP Steel Plate Table 2 shows the test items conducted to evaluate fitness of the 7%Ni-TMCP steel plate and its weld for inner shell of LNG storage tanks. The evaluation program is set to be the same as a previous study in order to evaluate 40mm, 50mm and 55mm heavy thick 9%Ni steel plate 7 8). Test plate thicknesses were 10, 25 and 40 mm, which were manufactured in actual production

4 Table 2 Evaluation program. Table 3 Chemical compositions [mass%]. Table 4 Test results of tensile test and Charpy test. equipment reflecting those findings mentioned in Chapter Results of Basic Mechanical Tests on Base Metal Table 3 shows results of chemical composition analysis. Impurity elements are kept to be very low by the latest steel making technology. Table4 shows results of tensile tests and Charpy impact tests. Both of them met the requirements of ASTM A553M-Type 1. Figure 7 shows 0.2% proof strength and tensile strength in low temperature tensile tests. Difference between 1/4t position and 1/2t position is scarcely recognized. Figure 8 shows strain-aged Charpy test results. Absorbed energy after straining remains in high to show that the degradation on absorbed energy by even though 5 percent straining is limited. 3.2 Results of Fracture Toughness Tests on Base Metal CTOD tests 12) were performed for evaluation of resistance to Fig. 7 Low temperature tensile test results.

5 134 Kubo etc. Development of 7%Ni-TMCP Steel Plate for LNG Storage tanks Fig. 9 Duplex ESSO specimen. Fig. 8 Strain aged Charpy test results. Table 5 CTOD test results. Table 6 DT test results. Fig. 10 Example of fracture surface of duplex ESSO test (25mmt,- 196deg. C). Table 7 Duplex ESSO test results. brittle crack initiation. As shown in Table 5, the fracture types always keep being "6" in every thickness and direction, which indicates no brittle fracture. Dynamic tear (DT) tests compliant with ASTM E604 15) were performed for estimation of brittle crack arrest toughness 11). The DT absorbed energy at -196 deg. C shows more than 2300J and fracture surface is almost covered with ductile manner as shown in Table 6. This indicates quite good property of brittle crack propagation. Furthermore, Duplex ESSO tests were performed under loading 393 MPa at -196deg.C to directly evaluate the arrest toughness (Figs. 9 and 10 and Table 7). Results of both 25mm and 40mm thick plate specimens showed "No-Go". 3.3 Results of Basic Mechanical Tests on Welded Joints Welded joints of 7%Ni-TMCP steel plates were prepared in three different welding methods; SMAW, GTAW and SAW (Table 8). V groove in case of 10mm thick plate and double-v groove in case of 25, 40mm were applied for edge preparation after actual tank fabrication. 70 percent nickel type welding consumables which are applied in actual welded joint of 9%Ni

6 steel plate were used. Figure 11 shows tensile strength of weldment tested by JIS1A specimen. In every thickness and welding condition, tensile strengths show more than 685MPa that is specified Table 8 Welding conditions. minimum value in A553M-Type1. Fractured appearances of tensile test of weldment and Vickers hardness in case of 25mm are exemplified in Fig.12. Typical hardness distributions of under-matching weldment 16 17), which causes advantage of prevention of brittle fracture, due to application of high Ni welding consumable are shown. Harder weld metals are SMAW, GTAW and SAW, in that order. In case of SMAW, fractured positions are located at base plate, so tensile strength of SMAW joints should be almost same as base plate. Table 9 shows Charpy impact test results at fusion line (FL) Fig. 11 Tensile test results of welded joints. Fig. 12 Appearance of fractured specimen of tensile test of welded joints and Vickers hardness distributions (25mm). Table 9 Charpy test result of welded joints.

7 136 Kubo etc. Development of 7%Ni-TMCP Steel Plate for LNG Storage tanks and HAZ1mm. All of absorbed energy shows sufficiently high. Average value of absorbed energy in 3/4 sub-size specimen shows more than 55J and average value of that in full-size specimen shows 80J. Figure 13 shows an example of the results of longitudinally bending test, which is performed for the detection of minimal cracks in weld metal. These tests are performed by 10mm-thick specimen extracted from top and bottom surface of weldments of all thickness and welding conditions. No crack is observed in all specimens. Fig. 13 Example of longitudinal bending tests (40mm-GTAW). 3.4 Results of Fracture Toughness Tests on Welded Joints CTOD tests 12) were performed on three different thicknesses i.e. 10, 25 and 40 mm and three different welding methods that were SMAW, GTAW and SAW. Table 10 shows results of weldment of 40mm thick plate that is maximum thickness, as an example. The critical CTOD value at -165 deg. C shows 0.4mm in minimum. The results of 10mm and 25mm thick plate are comparable. Furthermore, cross weld notch wide plate tests with a 2- times-thickness-long artificial through thickness notch were performed on 25 and 40 mm thick test specimens both for GTAW and SMAW joints that are vertical weldment in actual tank and normal to principal stress, to evaluate safety of the T-cross welded joint in LNG storage tanks (Figs. 14 and 15). The notch is located at fusion line (FL) and radius of tip is sharpened to 0.1mm by electro discharge. Test temperature is set to be 165 deg. C that is design temperature of LNG tank. The tests resulted in that fracture net stresses of all specimens showed more than 750MPa and much higher values than 400MPa, design stress based on huge scenario earthquake. Also CTOD values calculated by BCSS model 18) are quite high. All specimens were fractured in a ductile manner and no brittle pop-in was observed. Cracks initiated at the HAZ deviated into weld metal with no exception in Table 10 CTOD test results of welded joints (40mm,-165deg. C). Table 11 Test results of cross weld notch wide plate tensile test.

8 Fig. 15 Detail of notch and notch position. Fig. 14 Specimen of cross weld notch wide plate test. Fig. 17 Enlarged photo of fracture surface near the notch tip (25mmt, GTAW, fracture surface). Fig. 16 Fracture path of cross weld notch wide plate test (25mmt, GTAW, fracture path). all specimens (Figs. 16 and 17). Considering of fracture stress, it is concluded that initiation and propagation of fracture occur almost after yielding of the ligament section and so high safety against fracture is shown. 4. Comparison with properties to existing 9%Ni steel plate As shown in the preceding chapter, properties of the 7%Ni- TMCP steel plate are widely evaluated. In this chapter, safety against fracture of the 7%Ni-TMCP steel plate is evaluated comparing with the properties of the conventional 9%Ni steel plate. 4.1 Brittle Crack Initiation Properties Figure 18 shows comparison of the critical CTOD of the base plate between the 7%Ni-TMCP and 9%Ni steel plates. The critical CTOD of the thinner plate shows smaller values than the thicker plates. All test specimens showed fracture type "6". So it does not reflect a low toughness level in thinner plate but seems to Fig. 18 Comparison of CTOD test results of base plate between 7%Ni- TMCP and 9%Ni steel. be attributed to a smaller absolute value of ligament of specimen 19) and this does not reflect low toughness level in the thinner plate. It is also not considered that there is an inferiority in the thinner plate, as all values are scattered within the same range of the 9%Ni steel plate. Figures. 19(a) to 19(c) shows comparison of Charpy impact properties of welds. Absorbed energies of the 7%Ni-TMCP steel

9 138 Kubo etc. Development of 7%Ni-TMCP Steel Plate for LNG Storage tanks plate are comparable to the 9 %Ni steel plate in all notch locations and welding conditions. Comparison of the critical CTOD values of welds is shown in Figs.20 (a) to 20(c) except for the fully ductile weld metal. All results including pop-ins are distributed within the range of the 9%Ni CTOD test results. Furthermore, comparison of the maximum net stresses of the cross weld notch tensile test is shown in Fig. 21. The net stresses of the 7%Ni-TMCP steel is considered comparable to that of the 9 Fig. 19 Comparison of Charpy test results of welded joint between 7%Ni-TMCP and 9%Ni steel. Fig. 20 Comparison of CTOD test results of welded joint between 7%Ni-TMCP and 9%Ni steel.

10 Fig. 21 Comparison of net stresses of cross weld notch wide plate test between 7%Ni-TMCP and 9%Ni steel. Fig. 22 Comparison of DT test results between 7%Ni-TMCP and 9%Ni steel. %Ni steel. As all specimens show a ductile fracture manner in those of austenitic weld metal a brittle fracture property cannot be evaluated. It can be said that the welds of the 7%Ni-TMCP steel has a quite high safety against brittle fracture as same as that of the 9 %Ni steel. 4.2 Brittle Crack Propagation Properties Table 12 Comparison of Duplex ESSO test result between 7%Ni- TMCP and 9%Ni steel. DT test is performed as simplified estimation method for propagation property. As shown in Fig. 22, absorbed energy of the 7%Ni-TMCP steel plate is much superior to that of the conventional QT type 9%Ni steel plate. Also, in Duplex ESSO test, "No-Go" is obtained under loading 393 MPa at 196 deg. C that is a more severe condition compared to the actual operation (Table12). The reason why the 7%Ni-TMCP steel plate has superior propagation property in spite of lowering nickel by 2 percent is thought to be effect of retained austenite produced by TMCP with intermediate heat treatment. 5. Conclusion Mechanical properties of the 7%Ni-TMCP steel plate in 10, 25 and 40 mm thicknesses were widely evaluated for application to the above-ground LNG storage tanks by conducting large scale mechanical tests as well as small scale mechanical tests. These tests indicated that the 7%Ni-TMCP steel plate had quite excellent properties against brittle fracture equivalent to the existing 9%Ni steel plate. Hence it is concluded that the 7%Ni-TMCP steel plate warrants serious consideration for use in LNG storage tanks. A future study plans to investigate the 50 mm thick 7%Ni- TMCP steel plate by conducting a series of mechanical tests to investigate fracture toughness and by assessing safety of the large capacity LNG storage tank under seismic behaviours. References 1) G.R. Brophy and A.J. Miller: The Metallography and heat treatment of 8 to 10% Nickel steel, Transactions of the A.S.M.,Vol.41, pp (1948). 2) T. Kawabata, K. Onishi and A.Inami: Steel plate for low temperature service storage tank, Sumitomo metals, 50, pp (1998). (in Japanese) 3) M. Hoshino, N. Saitou, H. Muraoka and O. Saeki: Development of Super-9%Ni Steel Plates with Superior Low - Temperature Toughness for LNG Storage Tanks, Nippon Steel Technical Report, 380, pp ( 2004). (in Japanese) 4) T. Kubo, A. Omori and O. Tanigawa: Properties of High Toughness 9% Ni Heavy Section Steel Plate and Its Applicability to kl LNG Storage Tank, KAWASAKI STEEL GIHO, 30, pp (1998). (in Japanese) 5) N. Katayama: The-State-of-the-Art and the Subjects of Materials for LNG Storage Tanks and Piping, Journal of the Japan Welding Society, 73 (7), pp (2004). (in Japanese) 6) T. Kawabata, K. Fujiwara, K. Arimochi and H. Hirose: Development of 7%Ni-TMCP Steel Plate for LNG Storage tanks, Pre-prints of Autumn Meeting of HPI, pp (2005). (in Japanese) 7) S. Machida, N. Ishikura, N. Kubo, N. Katayama, S. Muramoto, Y. Hagiwara and K. Arimochi: Fracture Characteristics of Heavy Thickness 9% Ni Steel Plate and its Applicability to Large Scale

11 140 Kubo etc. Development of 7%Ni-TMCP Steel Plate for LNG Storage tanks LNG Storage Tanks (2nd Report, High Toughness 50-55mm Thick 9% Ni Steel Plate), JHPI, Vol.29, No.6, pp (1991). (in Japanese) 8) S. Machida, N. Ishikura, N. Kubo, N. Katayama, Y. Hagiwara and K. Arimochi: Fracture Characteristics of Heavy Thickness 9% Ni Steel Plate and its Applicability to Large Scale LNG Storage Tanks, JHPI, Vol.31, No.1, pp (1993). (in Japanese) 9) Sumitomo Metal Industries, Ltd., Internal Report, (2002). 10) Sumitomo Metal Industries, Ltd., Production Record, (2007~2008). 11) Consortium of Five Japanese Companies (OG, IHI, TKK, NSC, SMI): Crack Arrest Properties of 9% Ni Steel and Relation Between Crack-Initiation and Crack Arrest Tests, Gas Research Institute, Report GRI-86/0007, (1986). 12) British Standard BS7448 part1: Fracture Mechanics toughness tests Part1. Method for Determination of K IC, critical CTOD and critical J values of metallic materials, (1991). 13) T. Hasegawa, H. Mabuchi and S. Aihara: Influence of Chemical Compositions on Toughness of Steels with the Microstructure of As-quenched Martensite, Tetsu- to- Hagane, 82 (7), pp (1996). (in Japanese) 14) N. Saitou, A. Toyofuku and S. Yano: Development of 9%Ni Steel with High Brittle Crack Arrest Capability by Direct Quenching Process, Seitetsu-Kenkyu, 318, pp (1985). (in Japanese) 15) American Society for Testing and Materials, ASTM E604: Standard Test Method for Dynamic Tear Testing of Metallic Materials, 1983 (Reapproved 1994). 16) K. Arimochi, M. Nakanishi, S. Satou, F. Minami, M. Toyoda and K Satou: A Proposal on 3 Point Bending COD Test Method for Welded Joint of 9% Ni Steel, Journal of the Japan Welding Society, 52 (2), pp (1983). (in Japanese) 17) K Satou, M. Toyoda, F. Minami, S. Satou, M. Nakanishi and K. Arimochi: Crack Tip Plastic Deformation of Notched Plates with Mechanical Heterogeneity, Journal of the Japan Welding Society, 52 (2), pp (1983). (in Japanese) 18) B.A. Bilby, A.H. Cottrell, E. Smith and K.H. Swinden: Plastic Yielding from Sharp Notches, Proc.Royal Soc. A279, pp1 9 (1964). 19) T. Hasegawa, T. Haze, S. Aihara, K. Okamoto: Ductile Fracture Bchavior in CTOD Test of Multi-pass Weld Heat Affected Zone of Quenched and Tempered High Strength Steels, Tetsu- to- Hagane, vol.81, No.5, pp (1995). (in Japanese)