ECS Transactions, 3 (31) 89-97 (2007) 10.1149/1.2789218, copyright The Electrochemical Society

Influence of Cold Working on the Resistance of Stainless Steels to Pitting Corrosion L. Pegueta, b, B. Malkia, and B. Barouxa a

Institut National Polytechnique de Grenoble, ENSEEG/LTPCM, 1130, rue de la piscine, Saint Martin d’Hères 38402, France b UGINE&ALZ Research Center (Arcelor), rue Roger Salengro, Isbergues 62330, France The influence of cold rolling and tensile deformation on the resistance to pitting corrosion of two stainless steels (AISI 304 and AISI 430) has been investigated. Two electrochemical techniques under free corrosion and potentiostatic conditions have been carried out so that pit initiation and propagation could be analysed separately. Plastic deformation is shown to act differently depending on the pitting stage under consideration. (i) The pitinitiation frequency shows a maximum after 20% cold rolling reduction or 10% tensile deformation. (ii) The pit propagation rate increases monotonously with cold rolling reduction. In addition, it is evidenced that strain-induced martensite is not a necessary condition for the susceptibility of cold worked SS to pit initiation. The unexpected behaviour of pitting initiation is more likely related to the dislocations features resulting from the successive cold-working stages Introduction

Pitting corrosion of stainless steels (SS) is usually regarded as a complex mechanism involving electrochemical and metallurgical factors (1). The metallurgical aspect of the question is worth taking into account especially when dealing with industrial steels. Hence, the aim of this paper is to focus on the particular point of cold working. The metallurgical structure is deeply affected by cold working: This is the creation and the slip of dislocations which lead to plastic deformation of metals (2). Furthermore, austenitic grades can be sensitive to martensitic transformation at room temperature (3). Lastly, plastic deformation may cause inclusion elongation or fractures at the interface with the matrix (4, 5, 6). Only few results are available on the localized corrosion of cold worked SS. In addition, they are almost exclusively based on the effect of plastic deformation on the pitting potential (Vpit). Those particularly scattered data do not lead to practically reliable conclusions. Indeed Vpit has been reported either to decrease (7, 8, 9, 10, 11, 12, 13, 14, 15) to increase (7, 16, 17, 18) to be non affected (8, 10) and even to be non monotonously modified (19, 20, 21) by plastic deformation. Regarding the propagation current density of pits, it is generally found to increase as a function of cold working (17, 22) but not in every case (22). The present study proposes an approach targeting on AISI 304 and AISI 430 SS cold worked by rolling or tensile test and using different electrochemical techniques supplemented with metallurgical characterizations. Such an approach makes it possible to prospect independently the influence of cold working on the initiation and propagation stages of the pitting process and has proved its efficiency (23).

89

ECS Transactions, 3 (31) 89-97 (2007)

Experimental Materials One experimental austenitic stainless steel produced in the laboratory and two commercial grades (Arcelor) were studied: (i) AISI 304 (Lab.) prepared in laboratory had undergone a cold-rolling deformation process. (ii) AISI 304 and AISI 430 commercial grades were scheduled for tensile testing. Their chemical analysis is listed in Table I. TABLE I. Chemical composition of the investigated materials in weight %.

Grade

Elaboration

%C

% Mn

% Ni

% Cr

% Mo

% Cu

AISI 304 AISI 304 AISI 430

Laboratory Industrial Industrial

0.026 0.037 0.042

1.45 1.42 0.38

8.57 8.66 0.16

17.86 18.18 16.26

0.20 0.25 0.03

0.20 0.22 0.04

S (ppm) 51 12 21

%N 0.036 0.038 0.027

AISI 304 (Lab.) grade was cast as a 25 kg ingot under argon atmosphere. Next, plates were obtained by hot and cold rolling with intermediate annealing, leading to 10%, 20%, 30%, and 70% of final cold thickness reduction. Disks with a diameter of 15 mm were stamped from annealed and cold-rolled plates, to be investigated by electrochemical measurements. The induced α’-martensite content (Figure 1) as well as mechanical properties including Vickers hardness and yield stress show a monotonous increase with cold-rolling reduction. 60

martensite (%)

50 40 30 20 10 0 annealed

10%

20%

30%

70%

cold rolling reduction (%)

Figure 1. α’-martensite fraction in AISI 304 (Lab.) grade as a function of cold rolling reduction. Neither modifications of inclusions morphology nor evidence of decohesion between inclusions and matrix were observed by Scanning Electronic Microscopy. On a smaller scale, dislocations structure arrangements analyzed by Transmission Electronic Microscopy (TEM) are representative of the transition through each cold-working stage (24, 25). Thus a planar structure favouring dislocations pile-ups is seen at 20% of cold

90

ECS Transactions, 3 (31) 89-97 (2007)

reduction, while at 70% the dislocations network evolves toward a cellular structure typical of a dynamic recovery stage (Figure 2).

1 µm

(a)

500 nm

200 nm

(b)

(c)

Figure 2. TEM Images (white field) of AISI 304 (Lab.) grade (a) annealed, (b) 20% cold rolled and (c) 70% cold rolled. Tensile strain was applied to 275 mm×25 mm test specimens cut from commercial grades plates 1 mm thick, from which 15 mm diameter disks were stamped for electrochemical study (Figure 3). Engineering strains of 10%, 35%, and 60% for AISI 304 grade, and 10% and 20% for AISI 430 grade, were chosen, to take into account the lower total elongation of the ferritic grade (25%) as compared to the austenitic one (70%). Disks stamped from cold-rolled and strained SS were prepared following the same procedure: specimens were polished with SiC paper to a 1200 grit finish and afterwards with diamond paste up to 3 µm grade. They were degreased in an ultrasonic acetone/ethanol mixed bath, rinsed with distilled water, dried, and finally aged for 24 hr in ambient air. (a) (b) (c)

(d)

Figure 3. Image of AISI 304 samples before (a) and after tensile test up to 10% (b), 35% (c) and 60% (d) elongation (strained samples are stamped).

91

ECS Transactions, 3 (31) 89-97 (2007)

Electrochemical measurements Pitting potential (Vpit) is known to include in a single criterion the different stages of the pitting process (24). Accordingly, two other selected electrochemical techniques were employed, so that the pitting steps (i.e., pit initiation and propagation) could be studied independently. Pitting Transients at Open Circuit Potential. Pitting transients measurement is carried out using a 3-electrode setup operating at open circuit potential (26, 27). This system includes two identical electrodes connected through a zero resistance ammeter (ZRA) and a reference electrode (SCE), allowing simultaneous measurement of both intensity and potential transients in aerated (0.1 M NaCl + 2.10-4 M FeCl3) or (1 M NaCl + 10-3 M FeCl3). The variation of current signal with time was measured over 24 hr periods. The typical signal shows a succession of current fluctuations or transients. Each of these was considered to be the signature of the appearance, propagation, and repassivation of a metastable pit (28, 29, 30) (i.e., a pit that has repassivated after a few seconds of propagation). A sampling frequency of 18.75 Hz was selected, to provide sufficient resolution for these events to be studied. “PPR” test under potentiostatic conditions. “Pit Propagation Rate” test is a method first proposed by Syrett (31) in which the specimen is subjected to a potential cycle. In this adaptation, the potential is first potentiodynamically scanned in a deaerated 0.5 M NaCl, pH 6.6, 23°C electrolyte at a scan rate of 10 mV/min from OCP to a given potential between free potential Vcorr and Vpit. It is held at this potential for 10 min to obtain a steady-state of the current density in the passive condition. The potential scan is then continued to potentials more noble than Vpit until the current reaches a value of 10 mA, corresponding to the growth of numerous pits whose diameters may reach a few tens of µm (Figure 4). The potential is then decreased in a single step to a value between Vcorr and Vpit and held at this value for 10 min. Since no new stable pits appear at potentials below Vpit, the recorded current is a measure of the rate of existing pit growth. After the potential is left at free potential in order to repassivate the pits, the potential is scanned to the selected value once again to ensure that passivity is achieved. The average value of current during the 10-min pit-growth period is determined by graphic integration of the current vs. time recording. The total area actually pitted (total projected pit area) is determined by microscopic examination, and the pit propagation rate is calculated by dividing the average current by the area actually pitted.

Figure 4. SEM image showing an example of a macroscopic pit after « PPR » test on AISI 304 grade.

20µm

92

ECS Transactions, 3 (31) 89-97 (2007)

Results

Metastable pits counting The pitting corrosion initiation stage was studied by focusing on the cumulated number of transients appearing at Open Circuit Potential in NaCl 0.1 M + FeCl3 2.10-4 M (Figure 5). This figure increases with cold-rolling reduction. However, the number of metastable pits shows a maximum for 20% cold reduction, where it is four times higher than in the annealed state. 600 Cumulative number of transients

20% 500

30% 400

10% 300

70%

200

0% 100

0 0

20000

40000

60000

80000

time (s)

Figure 5. Cumulative number of transients during 24h on AISI 304 grade at free potential in NaCl 0.1M + FeCl3 2.10-4 M as a function of cold reduction. This rather surprising maximum in initiation frequency was confirmed by studying the influence of tensile deformation on both the AISI 304 and AISI 430 experimental grades. Indeed, a similar counting of metastable pitting transients at Open Circuit Potential during 24 hr in NaCl 1 M + FeCl3 10-3 M showed a significant rise in the initiation frequency for a 10% elongation (TABLE II). TABLE II. Number of metastable pits initiated during 24h on industrial AISI 304 and AISI 430 grades at

free potential in NaCl 1M + FeCl3 10-3 M as a function of strain. annealed 10% 20%

35%

60%

AISI 304

116

297

-

64

40

AISI 430

371

515

309

-

-

Estimated pit propagation rate The macroscopic propagation stage was investigated under applied potential by applying the PPR test to annealed, 20%, and 70% cold-rolling reductions. An expected

93

ECS Transactions, 3 (31) 89-97 (2007)

increase of the dissolution current density with applied potential is first evidenced. In addition a significant and monotonous rise with cold-rolling reduction is also found (Figure 6).

-2

current density (mA.cm )

160 0%

20%

70%

100

150

200

120

80

40

0 50

250

300

potential (mV/SCE) Figure 6. Propagation current density of macroscopic pits measured by « PPR » test on AISI 304 as a function of cold rolling reduction and applied potential. Discussion

The present study has demonstrated two distinct effects of cold working on SS localized corrosion, depending on the stage of the pitting process: the first is a nonmonotonous trend as a function of deformation while the second is fairly linear. In addition, strain-induced martensite is evidenced as a non-necessary condition for the susceptibility of cold worked SS to pit initiation. At the initiation stage, the main unexpected - indeed, relatively rare - behaviour (19, 20, 21) is the non-monotonous dependence on cold-working reduction. A maximum of metastable pits number is found at 20% of cold-rolling reduction or 10% of tensile deformation. This behaviour eliminates any correlation with the linear evolution of macroscopic metallurgical parameters such as yield stress or hardness. An understanding of the mechanisms involved would require a smaller-scale analysis owing to the fact that in stainless steels, pit initiation is closely related to non-metallic inclusions acting as triggers for future pits. The shapes of inclusion are likely to affect the pitting (32,33), and cracks in the inclusions or at the inclusion-matrix boundary (4, 5,6) are thought to be detrimental to initiation (11). Nevertheless, such morphological modifications of inclusions have not been found in the present work. TEM analysis suggests that it may rather be related to dislocations features produced during the various cold-working stages. This interpretation is supported by TEM images showing planar deformation structures for 20% cold rolling while development of cellular structures is observed at higher

94

ECS Transactions, 3 (31) 89-97 (2007)

reductions (Figure 2). Multiplication of dislocations as well as their arrangement into pile-ups induce high stress concentrations (2) and are likely to modify the local deformation potential (34). A mechano-electrochemical model has even been developed, indicating that a maximum of dislocations pile-ups is likely to increase the potential difference between inclusion and surrounding matrix (35), such local electrochemical heterogeneity is expected to enhance pit initiation. Moreover, some workers have noted that the specific combination of MnS inclusion and applied mechanical stress could affect the susceptibility to pitting (36). For the highest cold-rolling reductions, the dynamic recovery process of dislocations followed by rearrangement into cellular structures (25,26) could limit the previous effect and cause the decrease observed in the initiation frequency. The pitting initiation sensitization, seen even in ferritic grades, contradicts the suggestion that strain-induced martensite is the main factor governing this preliminary corrosion step. Many authors acknowledge that the absolute amount of martensite has very little importance (7, 11, 20, 37, 38, 39,40) and can hardly be separated from the influence of dislocations and internal stresses. Considering that martensitic transformation proceeds by oriented rapid dislocations movements and multiplication (41), a possible indirect effect of martensite cannot be ruled out to the extent that it favours planar structures and the stability of dislocations pile-ups. At a macroscopic scale, the pit propagation rate is found to increase monotonously with cold working. We suggest that this is related more to the total density of dislocations induced by cold working (macroscopic effect) than to local structures such as pile-ups (mesoscopic effect). Plastic deformation is believed to increase the density of dislocations, which in turn enhances dissolution, owing to the presence of lower bonding energy points compared to “perfect” crystals (42). Such experimental evidence provides new insights into the validity of thermodynamic approaches for assessing the dissolution rates in metals containing structural defects (35, 36, 43). Conclusions

From the results of the present investigation, plastic deformation was evidenced to act differently depending on the pitting stage under consideration. The following conclusions can be made: (i) At the initiation stage, a maximum of metastable pits is found for 10% tensile deformation or 20% cold-rolling reduction. It is suggested, based on a mechano-electrochemical approach, that dislocations features resulting from the successive cold-working stages explain this trend. (ii) This behaviour, seen even in ferritic grades, contradicts the hypothesis that strain-induced martensite is the main factor governing sensitivity to corrosion. (iii) The dissolution rate increases monotonously at the macroscopic propagation stage, when the overall density of dislocations is more probably the relevant factor. References

1. B. Baroux, in Further insights on the pitting corrosion of stainless steels, Ph. Marcus (Ed.), Corrosion Mechanisms in Theory and Practice, Marcel Dekker, New York, Chapter 10, pp. 311-347 (2002).

95

ECS Transactions, 3 (31) 89-97 (2007)

2. J. Friedel, Dislocations, Pergamon Press (1964). 3. S. S. Hecker, M. G. Stout, K. P. Staudhammer, J. L. Smith, Metallurgical Transactions, 13A, 619 (1982). 4. S. Rudnik, Journal of the Iron and Steel Institute, 374 (1966). 5. E. Rozvsky, W.C. Hahn, Jr., B. Avitzur, Metallurgical Transactions, 4, 927 (1973). 6. V. Vignal, R. Oltra, C. Josse, Scripta Materialia, 49, 779 (2003). 7. A. Barbucci, G. Cerisola, P. L. Cabot, Journal of the Electrochemical Society, 149, B534 (2002). 8. D. Sinigaglia, B. Vicentini, G. Taccani, G. Salvago, G. Dallaspezia, Journal of the Electrochemical Society, 130, 991 (1983). 9. G. Salvago, G. Fumagalli, D. Sinigaglia, Corrosion Science, 23, 515 (1983). 10. H. J. Torwie, S. Krutz, R. Bauer, W. Schaarwächter, Int. Congr. Met. Corros., 8th, 92 (1981). 11. B. Mazza, P. Pedeferri, D. Sinigaglia, A. Cigada, G. A. Mondora, G. Re, G. Taccani, D. Wenger, Journal of the Electrochemical Society, 126, 2075 (1979). 12. B. Mazza, P. Pedeferri, D. Sinigaglia, A. Cigada, A. Fumagalli, G. Re, Corrosion Science, 19, 907 (1979). 13. C. J. Semino, P. Pedeferri, G. T. Burstein, T. P. Hoar, Corrosion Science, 19, 1069 (1979). 14. A. Cigada, B. Mazza, P. Pedeferri, D. Sinigaglia, J.Biomed. Mater. Res., 11, 503 (1977). 15. B. Mazza, P. Pedeferri, D. Sinigaglia, A. Cigada, L. Lazzari, G. Re, D. Wenger, Journal of the Electrochemical Society, 123, 1157 (1976). 16. S.V. Phadnis, A.K. Satpati, K.P. Muthe, J.C. Vyas, R.I. Sudaresan, Corrosion Science, 45, 2467 (2003). 17. B. C. Syrett, S. S. Wing, Corrosion, 34, 138 (1978). 18. A. Baghdasarian, S. F. Ravitz, Corrosion, 31, 182 (1975). 19. B. Ravi Kumar, Raghuvir Singh, Bhupeshwar Mahato, P.K. De, N.R. Bandyopadhyay and D.K. Bhattacharya, Materials Characterization, 54, 141 (2005). 20. U. Kamachi Mudali, P. Shankar, S. Ningshen, R. K. Dayal, H. S. Khatak, Baldev Raj, Corrosion Science, 44, 2183 (2002). 21. V. A. C. Haanappel, M. F. Stroosnijder, Corrosion, 57, 557 (2001). 22. R. Stefec, F. Franz, Corrosion Science, 18, 161 (1978). 23. T. Sourisseau, E. Chauveau, B. Baroux, Corrosion Science, 47, 1097 (2005). 24. U. F. Kocks, H. Mecking, Progress in Materials Science, 48, 171 (2003). 25. E. Nes, Progress in Materials Science, 41, 129 (1998). 26. J. L. Dawson, D. A. Eden, R. N. Carr, Patent number PCT/GB92/00527, 23/02/92, Publication number WO 92/16825. 27. G. Berthomé, B. Malki, B. Baroux, Corrosion Science, 48, 2432 (2006). 28. Frankel et al., Corrosion, 43, 429 (1973). 29. N. Sato, Journal of the Electrochemical Society, 129, 260 (1982). 30. H. Isaacs, Corrosion Science, 29, 313 (1989). 31. B. C. Syrett, Corrosion, 33, 221 (1977). 32. P. Poyet, A. Desestret, H. Coriou, L. Grall, Mem. Sci. Rev. Met., 72, 133 (1975). 33. R. Ke, R. Alkire, Journal of the Electrochemical Society, 142, 4056 (1995). 34. B.I. Kolodii, Materials Science, 36, 884 (2000).

96

ECS Transactions, 3 (31) 89-97 (2007)

35. E.M. Gutman, Mechanochemistry of Solid Surface, World Scientific Publishing (1994). 36. T. Suter, E.G. Webb, H. Böhni, R.C. Alkire, Journal of the Electrochemical Society, 148, B174 (2001). 37. A. Barbucci, M. Delucchi, M. Panizza, M. Sacco, G. Cerisola, J. Alloys Compd., 317-318, 607 (2001). 38. A. Cigada, B. Mazza, G. A. Mondora, P. Pedeferri, G. Re, D. Sinigaglia, Corrosion and degradation of implant materials, ASTM STP 684, Philadelphia, 144 (1979). 39. B. Mazza, P. Pedeferri, D. Sinigaglia, A. Cigada, L. Lazzari, A. Mondora, G. Re, G. Taccani, D. Wenger, Proc. 7th Int. Congr. Metallic Corrosion, ABRACO, Rio de Janeiro, 484 (1979). 40. E.M. Gutman, G. Solovioff, D. Eliezer, Corrosion Science, 38, 1141 (1996). 41. X.F. Fang, W. Dahl, Materials Science and Engineering, A141, 189 (1991). 42. N.D. Greene, G.A. Saltzman, Corrosion, 20, 294t (1964). 43. Li De-Lin, Zhu Ri-Zhang, Zhang Wen-Qi, Science in China (Series A), 32(11), 1371 (1989).

97