CC BY-NC-ND
0Available online at www.sciencedirect.com
SciVerse ScienceDirect Procedia
Engineering
Procedia Engineering 55 (2013) 64 - 69 =
www. el sevi er. com/1 ocate/procedi a
6th International Conference on Creep, Fatigue and Creep-Fatigue Interaction [CF-6]
Creep Deformation and Rupture Behaviour of P92 Steel at 923 K
E. Isaac Samuel, B. K. Choudhary*, D.P. Rao Palaparti, M.D. Mathew
Mechanical Metallurgy Division, Indira Gandhi Centre for Atomic Research, Kalpakkam - 603102, Tamil Nadu, India
Abstract
Creep deformation and rupture behaviour of P92 steel has been examined at 923 K for stresses ranging from 75 to 150 MPa. The steel exhibited well defined primary, secondary characterized by minimum creep rate and prolonged tertiary creep stages. The stress dependence of minimum creep rate obeyed Norton's power law and exhibited distinct stress regimes characterised by separate values of stress exponents in low and high stress regimes. Similarly, the stress dependence of rupture life also obeyed power law and displayed two stress regimes with separate stress exponent values. The steel displayed decrease in creep ductility with increase in rupture life in the low stress regime and followed generalised Monkman-Grant relation interrelating minimum creep rate and rupture life. Modified Monkman-Grant relation has been found to be valid for the steel. Fractographic examination indicated dominance of transgranular fracture on the fracture surfaces of tested specimens.
© 2011 The Authors.Publi shedby Elsevier Ltd.
Selection and peer-review underresponsibility of the Indira Gandhi Centre for Atomic Research. Keywords: P92 steel; creep deformation; creep-rupture; creep ductility
1. Introduction
The chromium-molybdenum (Cr-Mo) family of ferritic steels form an important group of alloys developed as structural materials for elevated temperature applications in the chemical, petrochemical and fossil-fired power generating industries [1,2]. The properties of Cr-Mo steels are controlled by the microstructures such as type of structures, grain size, solid solution strengthening and precipitation hardening. Advancement towards new generation power plants such as Ultra Super Critical (USC) power plants has increased operating steam temperatures and pressures to achieve higher efficiency and better environmental protection. This has led to the development of improved versions of 9% Cr steels with excellent combination of creep strength and ductility for steam generator applications. The improved versions mainly include 9Cr-1Mo steel modified by the addition of strong carbide forming elements such Nb and V known as P91 steel [3]. The steel is further modified by the addition of W with reduced Mo designated as P92 steel [3]. As mentioned earlier, the driving force in the development of P91 and P92 steels has been to achieve improved creep-rupture strength, it is therefore essential to emphasize that understanding of creep behaviour is of paramount importance. Among the
* Corresponding Author: E-mail address: bkc@igcar.gov.in
ELSEVIER
1877-7058 © 2013 The Authors. Published by Elsevier Ltd.
Selection and peer-review under responsibility of the Indira Gandhi Centre for Atomic Research. doi: 10. 1016/j .proeng .2013.03.220
9% Cr family of steels, P92 steel exhibits superior creep-rupture strength at elevated temperatures [4-10]. In this paper, creep-rupture properties evaluated for a wide stress range at 923 K have been presented. Creep-rupture properties have been examined in terms of stress dependence of creep rate and rupture life, variations in creep ductility with rupture life and fracture behaviour. The applicability of Monkman-Gant (MGR) [11] and modified Monkman-Grant (MMGR) [12] relations to P92 steel, interrelating creep rate and rupture life useful for creep life extrapolation has been discussed.
2. Experimental
P92 steel pipe of dimension 260 mm outer diameter and 25 mm wall thickness was used in this study. The chemical composition of the steel was conforming to ASTM standards [3]. Specimen blanks of size 20 mm diameter and 130 mm length were machined in the length direction of the pipe. Specimen blanks were subjected to normalising at 1338 K for 2 h followed by air cooling and tempering at 1053 K for 2 h followed by air cooling. Microstructure in normalised and tempered (N+T) condition was composed of tempered lath martensite and precipitates along the lath boundaries and prior austenite grain boundaries. The intralath matrix regions contained fine precipitates. Creep specimens of dimension 50 mm gauge length and 10 mm gauge diameter were machined from the N+T specimen blanks. Uniaxial creep tests were carried out at 923 K employing stresses ranging from 75 to 150 MPa. The test temperature was maintained within ± 2 K during all the tests. The elongation values were measured using LVDT transducer at room temperature attached to the bottom of the high temperature extensometer mounted at the gauge end portion of the specimens. The strain resolution was 4x10-4 for LVDT transducer. Fractographic examinations were performed on creep tested specimens using scanning electron microscope (SEM).
3. Results and discussion
The variations of creep strain with time at different stress levels at 923 K are shown in Fig. 1. Creep deformation of P92 steel was characterised by a small amount of instantaneous strain followed by well defined but short primary, secondary and tertiary creep stages. Deceleration in creep rate from initial high value after loading can be seen in the double logarithmic plots of e vs. t in Fig. 2 at all stress levels. This decrease in creep rate in the primary creep approaches to a minimum creep rate, em, in the secondary creep. A systematic increase in the time to the occurrence of minimum creep rate with decrease in applied stress can be seen in Fig. 2. Following the occurrence of minimum creep rate, creep rate increases rapidly with time in tertiary creep resulting in failure. The variations in creep rate with creep strain are shown in the plots log e vs. loge in Fig. 3. A systematic decrease in creep rate in the primary creep followed by minimum creep rate and a prolonged tertiary creep in terms of large creep strain accumulation can be seen in Fig. 3. A systematic decrease in the strain to the occurrence of minimum creep rate from 3.2% at 150 MPa to 0.57% at 75 MPa has been observed (Fig. 3).
10'7 10
P92 Ferritic Steel V
- 923 K O
- O O B V
- V ^ftj^
□ 150 MPa
- O 137 MPa
A 125 MPa
- V 112 MPa
o 100 MPa
- < 87 MPa
> 75 MPa i 1
10 Time, h
Fig. 1. Creep curves obtained at 923 K for P92 steel at Fig. 2. Variations of creep rate with time obtained for different different stress levels. stress levels at 923 K for P92 steel.
P92 Ferritic Steel 923 K
<D 10"2
137 MPa 125 MPa 112 MPa 100 MPa 87 MPa 75 MPa
150 MPa t
0.01 0.1 1
Creep strain, %
Fig.3. Variations of creep rate with creep strain obtained for different stress levels at 923 K for P92 steel.
Stress dependence of minimum creep rate is shown as log £mvs. logo in Fig. 4. P92 steel obeyed Norton's power law ( s m = A o n , where A is constant and n is the power law exponent) exhibiting two-slope behaviour characterised by distinct values of stress exponent n as 4.4 and 10.2 in the low and high stress regimes, respectively. The variation of rupture life with applied stress is presented as log tr vs. logo in Fig. 5. Like, stress dependence of minimum creep rate, stress dependence of rupture life (tr) also obeyed power law of the form tr = A' G , where A' is constant and n' is the power law exponent. Two slope behaviour with separate values of exponent n' = 3.9 and 9.7 in the low and high stress regimes, has been observed. The variations in creep ductility measured in terms of % elongation (strain to failure) and % reduction in area with rupture life are shown in Figs. 6 and 7, respectively. Both % elongation and % reduction in area exhibited a marginal decrease in the high stress regime followed by rapid decrease in the values with increase in rupture life in the low stress regime. SEM investigations on the fracture surface of creep tested specimens revealed dominance of transgranular fracture resulting from coalescence of microvoids at all stress levels (Fig. 8). However, signatures of a few isolated intergranular cracks (shown by arrows in Fig. 8b) at the fracture surface in the low stress regime was observed.
Distinct stress regimes in the stress dependence of minimum creep rate and rupture life with separate respective values stress exponents observed in the present investigation is in agreement with those reported by Ennis et al. [9, 10], Kimura et al. [4,5] and Lee et al. [8] for P92 steel. Sklenicka et al. [7] reported high values of stress exponent i.e., n = 12 and 18, for the P91 and P92 steels, respectively, in the high stress regime at 873 K. In 9Cr-1Mo steel, Choudhary et al. [13] observed n = 5.5 and 10.2 in low and high stress regimes, respectively. Different values of stress exponent have been associated with different apparent activation energy values in the low and high stress regimes [8]. Invoking resisting stress concept, the different values of stress exponent and apparent activation energy in the two stress regimes have been rationalised into a single values of stress exponent and activation energy. Based on this, it was suggested that creep behaviour is controlled of climb of dislocations in both the stress regimes in 9Cr-1Mo steel [13,14]. Detailed analysis has been undertaken to resolve the observed two slope behaviour in P92 steel. Further, similar values of power law exponents n and n' observed for the stress dependence of creep rate and rupture life indicate that the creep deformation and fracture processes are not different but are same in P92 steel. The interrelation between minimum creep rate and rupture life is further demonstrated by the applicability of minimum creep rate-rupture life relations of Monkman-Gant [11] and modified Monkman-Grant [12] type for the steel described below.
Stress MPa Stress, MPa
Fig. 4. Variation of minimum creep rate with applied stress Fig. 5. Variation of rupture life with applied stress at 923 K for
at 923 K for P92 steel.
P92 steel.
Fig. 6. Variation in strain to failure with rupture life at 923 K Fig. 7. Variation in reduction in area with rupture life at 923 K for P92 steel. for P92 steel.
Fig. 8. Fracture surface of the P92 steel specimen after creep testing at 923 K at (a) 125 and (b) 75 MPa corresponding to rupture lives of 748 and 18,865 h, respectively. The arrow marks show the isolated evidences for intergranular cracking in (b).
-2 105
2 104 w
P92 Ferritic Steel
^sû 923 K
C 0.11
Minimum creep rate , h
Fig. 9. Rupture life vs. minimum creep rate plot showing applicability of generalised form of Monkman-Grant relation in P92 steel.
Fig. 10. Rupture life/strain to failure vs. minimum creep rate plot showing the validity of modified Monkman-Grant relation in P92 steel.
Figure 9 shows the variation in rupture life as a function of minimum creep rate. The steel followed generalised form of Monkman-Gant relation (MGR) expressed as
where a is slope of logtr vs. log em plot and C is constant. The values of a = 0.903 and C = 0.063 have obtained for P92 steel at 923 K. The observed value of a less than unity indicates that the contribution of secondary creep strain, i.e., £m • tr decreases with increase in rupture life and decrease in stress. The variation in tr/ef (where ef is strain to failure) with em depicting applicability of modified Monkman-Grant relation (MMGR) proposed by Dobes and Milicka [12] as
£a •
is shown in Fig. 10. In Eq. (2), a' and C' are the slope and intercept of log(tr/ef) vs. log em plot, respectively. The validity of modified Monkman-Grant relation in P92 steel is observed as the slope a' equals to unity in the plot of log(tr/ef) vs. log s m (Fig. 10) and Eq. (2) can be expressed as
£ m • — = constant = C £f
CMMG = 0.11 has been obtained for the steel. This suggests that the term (tr/ef) x è m do not vary with stress/rupture life at 923 K for the steel and CMMG is real constant independent of stress. It can be seen that contrary to stress dependence of creep rate and rupture life, both MGR and MMGR do not exhibit separate and different stress regimes. Further, the observed lower value of CMMG = 0.11 indicates that the contribution from secondary creep is small and the most of the creep strain is derived from tertiary creep in P92 steel. High creep damage tolerance factor X = 9 evaluated from X = 1/ CMMG indicates microstructural degradation as the dominant damage mechanism in P92 steel. Ennis et al. [9,10] observed microstructural degradation in terms of decrease in dislocation density, precipitate coarsening and subgrain coarsening in the steel. The microstructural degradation was more dominant at low stresses compared to that observed at high stresses. Choudhary et al. [14] reported similar influence of degradation in microstructure in the two stress regimes in 9Cr-1Mo steel. Apart from microstructural degradation in P92 steel, it is equally important to point out that the loss of creep ductility indicates occurrence of localised cracking particularly at longer durations in the low stress regime. The observed loss of ductility in the present study is in agreement with that reported by Lee et al. [8] for the steel.
Loss of creep ductility associated with intergranular cracking in the form of wedge cracks and r-type cavities in the low stress regime has been reported in P92 steel [8]. In view of this, detailed metallographic examination on longitudinally sectioned creep tested specimens have been undertaken to ascertain the creep damage mechanism appropriate for the steel.
4. Conclusions
Creep deformation of P92 steel indicated well defined primary, secondary characterised by minimum creep rate and prolonged tertiary creep stages. The stress dependence of minimum creep rate obeyed Norton's power law and exhibited two-slope behaviour with separate values of stress exponents. Similarly, the stress dependence of rupture life also obeyed power law and displayed distinct values of stress exponent in the low and high stress regimes. The steel displayed decrease in creep ductility with increase in rupture life in the low stress regime and followed generalised form of Monkman-Grant relation interrelating minimum creep rate and rupture life. Modified Monkman-Grant relation has been found to be valid for the steel. Apart from the dominance of microstructural degradation as shown by high creep damage tolerance factor, P92 steel also exhibited some evidence of damage due to localised cracking associated with loss of creep ductility at longer durations.
References
[1] R.L.Klueh, Elevated temperature ferritic and martensitic steels and their application to future nuclear reactors, Int. Mater. Rev., 50(2005)287.
[2] F.Masuyama, Creep rupture life and design factors for high-strength ferritic steels, Int. J. Press. Vessels and Piping, 84(2007)53-61.
[3] Standard Specification for Seamless Ferritic and Austenitic Alloy-Steel Boiler, Superheater, and Heat-Exchanger, Tubes. ASTM Standards 2011, A213/A213M - 11a.
[4] K.Kimura, K.Sawada, H.Kushima and K.Kubo, Effect of stress on the creep deformation of ASME Grade P92/T92 steels, Int. J. Mater. Res., 99(2008)395-401.
[5] K.Kimura, Y.Toda, H.Kushima and K.Sawada, Creep strength of high chromium steel with ferrite matrix, Int. J. Pres. Vessels and Piping, (2010)87; 282-88.
[6] K.Sawada, K.Kubo, F.Abe, Creep behavior and stability of MX precipitates at high temperature in 9Cr-0.5Mo-1.8W-VNb steel, Mater. Sci. Engg., (2001)A319-321; 784-7.
[7] V.Sklenicka, K.Kucharova, M.Svoboda, L.Kloc, J.Bursik, A.Kroupa, Long-term creep behavior of 9-12%Cr power plant steels, Mater. Charact., (2003)51; 35.
[8] J.S.Lee, H.G.Armaki, K.Maruyama, T.Muraki, H.Asahi, Causes of breakdown of creep strength in 9Cr-1.8W-0.5Mo-VNb steel. Mater. Sci. Engg., (2006) A 428; 270-5.
[9] P.J.Ennis, A.Zielinska-Lipiec, O.Wachter, A.Czyrska-Filemonowicz, Microstructural stability and creep rupture strength of the martensitic steel P92 for advanced power plant, Acta Mater., 45(1997)4901.
[10] P.J.Ennis, A.Czyrska-Filemonowicz, Recent advances in creep-resistant steels for power plant applications, Sadhana (2003)28; 70930.
[11] F.C.Monkman, N.J.Grant, An empirical relationship between rupture life and minimum creep rate in creep-rupture tests, Proc. Am. Soc. Test Mater., (1956)56; 593-620.
[12] F.Dobes, K.Milicka, The relation between minimum creep rate and time to fracture, Metal Sci., (1976)10; 382-4.
[13] B.K.Choudhary, K.B.S.Rao, S.L.Mannan, Steady state creep deformation behaviour of 9Cr-1Mo ferritic steel forging in quenched and tempered condition, Trans Indian Inst Met (1999)52; 327-36.
[14] B.K.Choudhary, S.Saroja, K.B.S.Rao, S.L.Mannan, Creep-rupture behaviour of forged, thick section 9Cr-1Mo ferritic steel, Metall. Mater. Trans., (1999)A30; 2825-34.
