Available online at Procedia Engineering 55 (2013 ) investigations.

Size: px
Start display at page:

Download "Available online at Procedia Engineering 55 (2013 ) investigations."

Transcription

1 Available online at Procedia Engineering 55 (2013 ) th International Conference on Creep, Fatigue and Creep-Fatigue Interaction [CF-6] Thermomechanical Fatigue Behaviour of a Modified 9Cr-1Mo Ferritic-martensitic Steel A. Nagesha a, R. Kannan a, R. Sandhya a, G.V.S. Sastry b, M.D. Mathew a, K. Bhanu Sankara Rao c, Vakil Singh b a Mechanical Metallurgy Division, Indira Gandhi Centre for Atomic Research, Kalpakkam, India b Centre of Advanced Study, Department of Metallurgical Engineering, Institute of Technology, Banaras Hindu University, Varanasi, India c School of Engineering Sciences and Technology, Central University, Hyderabad Abstract Thermomechanical fatigue (TMF) tests were carried out on a modified 9Cr-1Mo ferritic steel (P91) under a mechanical strain control mode using a strain amplitude of ±0.4% % and a strain rate of s -1. In-Phase (IP) and Out-of-Phase (OP) strain-time waveforms were employed for the tests which were performed under different temperature ranges in the interval, K. For the sake of comparison, isothermal LCF (designated as IF) tests were also carried out at the maximum temperatures (T max ) of TMF cycles on similar specimens and using the same strain amplitude and strain rate. Isothermal cycling was observed to be the most detrimental while IP TMF yielded the highest lives. However, with an increase in the T max of TMF cycling, the difference in lives was seen to narrow down. Also, lives under IP TMF and IF cycling were seen to reduce more drastically compared to OP cycling on account of a greater creep damage accumulation. A continuous cyclic softening characterized the stress response of the alloy under all testing conditions. The lower lives observed under OP cycling were rationalised in terms of oxidation damage and mean stress development. The observed behaviour was explained on the basis of detailed TEM investigations The Published Authors. Published by Elsevier by Elsevier Ltd. Selection Ltd. Open access and/or under peer-review CC BY-NC-ND under license. responsibility of the Indira Gandhi Centre for Selection Atomic Research. and peer-review under responsibility of the Indira Gandhi Centre for Atomic Research. Keywords: Thermomechanical fatigue; in-phase; out-of-phase; cyclic softening; oxidation; P91 steeel 1. Introduction Modified 9Cr-1Mo ferritic steel with alloying additions of niobium and vanadium, referred to as P91, is the chosen structural material for the steam generator components of liquid metal cooled fast breeder reactors. These components undergo thermally induced strain cycling during start-up and shut-down or during variation in the operating conditions, leading to the development of thermomechanical fatigue (TMF). Extensive literature exists on the isothermal low cycle fatigue (hereafter referred to as IF) behaviour of modified 9Cr-1Mo steel, both in air [1-8] and liquid metal [9-12] environments. However, reported literature on the TMF Corresponding author: address: nagesh@igcar.gov.in The Authors. Published by Elsevier Ltd. Open access under CC BY-NC-ND license. Selection and peer-review under responsibility of the Indira Gandhi Centre for Atomic Research. doi: /j.proeng

2 200 A. Nagesha et al. / Procedia Engineering 55 ( 2013 ) behaviour of the alloy is limited [13-15]. The present work aims at understanding the influence of strain temperature phasing on the fatigue deformation and damage behaviour of P91 steel under TMF cycling. Based on detailed TEM investigations, it was also attempted to compare the material behaviours under IF and TMF loading conditions. 2. Experimental In-Phase and Out-of-Phase TMF tests were carried out on tubular samples employing a mechanical strain amplitude (Δε mech ) of ± 0.4%. A fixed strain rate of s -1 was used for all the tests. Also, a constant temperature range (ΔT) of 200 K was used, and tests were carried out in the interval, K. Besides, IF tests were performed at the peak temperatures (T max ) of TMF cycling using the same strain rate and identical specimen geometry for a comparative evaluation of TMF and isothermal behaviours. Details pertaining to the machine and the experimental setup used are elaborately covered elsewhere [16]. The alloy had the following chemical composition (in wt. %): Cr: 9.3, C: 0.11, Mo: 0.99, Ni: 0.14, V: 0.25, Nb: 0.1, N: 0.068, S: and P: The material was given a normalizing (1313 K for 1 h + air cooling) followed by tempering (1033 K for 1 h + air cooling) treatment prior to testing. The microstructure of the heat-treated material comprised of a tempered martensitic lath structure with a high dislocation density. 3. Results and discussion 3.1. Cyclic stress response The alloy displayed a continuous cyclic softening under both TMF and IF tests, in all the temperatures ranges investigated. Figure 1 shows a comparative plot of the cyclic stress response obtained under TMF in the temperature interval of K and IF cycling at the T max (623 K). Such a behaviour is a characteristic feature of most ferritic-martensitic steels [1-15]. Upon cycling, the transformation of the lath into a subgrain structure took place (Figs. 2a-c), resulting in cyclic softening. It was also seen that IF cycling at the T max produced a significantly greater recovery compared to that induced by TMF cycling (Fig. 2c) owing to the sustained deformation taking place at the T max, compared to that resulted under TMF. As a consequence, IF cycling led to a lower stress response compared to TMF Peak Stress, MPa 200 IF test, 823 K IP TMF test, K OP TMF test, K Number of Cycles Fig. 1.Cyclic stress response plots under TMF ( K) and IF cycling at T max (823 K).

3 A. Nagesha et al. / Procedia Engineering 55 ( 2013 ) Influence of oxidation Ferritic steels are known to suffer considerable oxidation-assisted damage during high temperature thermal exposure [17]. Presence of external stresses, thermal or mechanical, leads either to cracks or detachment of oxide layer (spalling) that in turn exposes fresh material to the environment. Out-of-Phase TMF yielded consistently lower lives compared to IP cycling under all the temperature ranges investigated. This is consistent with the well established compressive dwell sensitivity of these class of alloys under creep-fatigue cycling reported by several investigators [2-5, 8]. Out-of-phase cycling may be considered similar to a compressive dwell creep-fatigue interaction test in that the plastic deformation and hence the creep damage predominates under compressive loading. Further, a tensile mean stress builds up under OP TMF just as it does, in the case of a compressive hold creep-fatigue interaction test. The base metal usually expands more than the oxide it forms during a heating cycle on account of its higher thermal expansion coefficient. Consequently, the oxide is subjected to a tensile stress that can only be relieved by cracking [18]. The tensile mean stress prevailing in OP cycling could also facilitate easy cracking of the oxide scale. The cracks under IP cycling were seen to be blunt and coated with thick oxide scales (Fig. 3a), as opposed to the relatively sharper cracks observed under OP TMF (Fig. 3b). Thus, the differences in lives between IP and OP cycling are likely to arise as a consequence of a higher propensity for oxide cracking under OP cycling. (a) (b) (c) Fig. 2. Substructures resulted under (a) IP TMF, (b) OP TMF and (c) IF cycling at 923K (ΔT for TMF: K).

4 202 A. Nagesha et al. / Procedia Engineering 55 ( 2013 ) μm (a) (b) Fig. 3. Secondary cracking observed under (a) IP and (b) OP TMF, K. Furthermore, OP TMF cycling was seen to be associated with consistently higher tensile hysteresis energy compared to IP cycling [15]. Also, the tensile mean stress under OP TMF (Fig. 2) ensured minimal contact between the mating crack surfaces, thereby resulting in rapid crack growth rates. Therefore, a combination of oxide cracking and a tensile mean stress contributed to lower lives under OP TMF. It should be noted that the base metal contracts more than its oxide during a cooling cycle leading to the development of a compressive stress in the oxide layer that is likely to get relaxed by the oxide spallation [18]. The compressive mean stress developed under IP TMF could also facilitate spallation of the oxide scale. In-Phase TMF cycling thus yielded better lives owing to a combined influence of crack tip blunting, together with a compressive mean stress. It should however be emphasized that the advantage associated with compressive mean stress under IP TMF gets offset on account of predominance of creep deformation as the maximum testing temperature increases. This is reflected in a more severe reduction in the half-life tensile stress with increasing temperature, seen under IF and IP TMF cycling conditions as compared to OP TMF, as presented in Fig. 4. Variation of cyclic life with T max indicated that life reduction was more drastic under IP TMF conditions in comparison with OP cycling, owing to the damaging contribution from creep deformation in the former case. Consequently, a progressive reduction in the difference between the lives obtained under TMF (IP, OP) and IF cycling with increasing T max was noted, as presented in Fig. 5. Fig. 4. Half-life tensile stress vs. temperature plots under TMF (IP, OP) and IF cycling at Tmax. Fig. 5. Comparative plots of the cyclic lives obtained under TMF (IP, OP) and IF cycling at Tmax [15].

5 A. Nagesha et al. / Procedia Engineering 55 ( 2013 ) Conclusions Isothermal LCF and TMF behaviours of modified 9Cr-1Mo ferritic steel were investigated. isothermal cycling at the maximum temperatures of TMF cycling yielded lowest lives compared to both IP and OP TMF cycling. Also, OP TMF cycling yielded lower lives compared to IP cycling under all the temperature ranges investigated. The lower lives observed under OP TMF in comparison with IP cycling were explained on the basis of oxidation effects coupled with a tensile mean stress in the former. Also, cyclic life under IP TMF cycling was observed to reduce more drastically compared to OP cycling, owing to the development of a greater amount of creep damage which gains significance with increasing T max. Consequently, the difference in lives under IF and TMF cycling narrowed down with an increase in the T max of TMF cycling. Acknowledgements The authors wish to acknowledge the encouragement received from Dr. T. Jayakumar, Director, MMG and Dr. A.K. Bhaduri, Associate Director, MDTG, IGCAR, Kalpakkam. Help received from Shri. G. Sukumaran and Shri. M. Srinivasa Rao, of Mechanical Metallurgy Division, IGCAR, Kalpakkam is gratefully acknowledged. References [1] A.Nagesha, M.Valsan, R.Kannan, K.Bhanu Sankara Rao and S.L.Mannan, Int. J. Fatigue 24(2002)1285. [2] B.Fournier, M.Sauzay, C.Caës, M.Noblecourt, M.Mottot, A.Bougault, V.Rabeau and A.Pineau, Int. J. Fatigue 30(2008)663. [3] B.Fournier, M.Sauzay, F.Barcelo, E.Rauch, A.Renault, T.Cozzika, L.Dupuy and A.Pineau, Metall. Trans. 40A(2009)330. [4] B.Fournier, F.Dalle, M.Sauzay, J.Longour, M.Salvi, C.Caës, I.Tournié, P.-F.Giroux and S.-H.Kim, Mater. Sci. Eng. A528(2011) [5] S.Kim and J.R.Weertman, Metall. Trans. A 19(1988)999. [6] W.B.Jones, in Ferritic Steels for High Temperature Applications, Proc. ASM International Conference on Production, Fabrication, Properties and Application of Ferritic Steels for High Temperature Applications, Warren, PA, October 6-8, 1981, Ed. A.K. Khare, ASM, Metals Park, Ohio, (1983) pp [7] G.Ebi and A.J.McEvily, Fatigue Fract. Eng. Mater. Struct. 17(1984)299. [8] Vani Shankar, M.Valsan, K.Bhanu Sankara Rao, R.Kannan, S.L.Mannan and S.D.Pathak, Mater. Sci. Eng. A 437(2006)413. [9] R.Kannan, R.Sandhya, V.Ganesan, M.Valsan and K.Bhanu Sankara Rao, J. Nucl. Mater. 384(2009)286. [10] R.Kannan, V.Ganesan, K.Mariappan, G.Sukumaran, R.Sandhya, M.D.Mathew and K.Bhanu Sankara Rao, Nucl. Eng. Des. 241(2011)2807. [11] J.-B. Vogt, A.Verleene, I.Serre and A.Legris, J. Nucl. Mater. 335(2004)222. [12] A.Weisenburger, A.Heinzel, C.Fazio, G.Müller, V.G.Markow and A.D.Kastanov, J. Nucl. Mater. 377(2008)261. [13] A.Marek, G.Junak and J.Okrajni, Arch. Mater. Sci. Eng. 40(2009)37. [14] Vani Shankar, V.Bauer, R.Sandhya, M.D.Mathew and H.-J.Christ, J. Nucl. Mater. 420(1 3)(2012)23. [15] A.Nagesha, R.Kannan, G.V.S.Sastry, R.Sandhya, Vakil Singh, K.Bhanu Sankara Rao and M.D.Mathew, Mater. Sci. Engg. A 554(2012)95. [16] A.Nagesha, M.Valsan, R.Kannan, K.Bhanu Sankara Rao, V.Bauer, H.-J.Christ and Vakil Singh, Int. J. Fatigue 31(2009)636. [17] K.D.Challenger, A.K.Miller and R.L.Longdon, J. Mater. Energy Systems 3(1981)51. [18] D.L.Deadmore and C.E.Lowell, Oxid. Met. 11(1977)91.