Failure analysis of cold drawn prestressing steel wires subjected to stress corrosion cracking

Failure analysis of cold drawn prestressing steel wires subjected to stress corrosion cracking

Engineering Failure Analysis 12 (2005) 654–661 www.elsevier.com/locate/engfailanal Failure analysis of cold drawn prestressing steel wires subjected ...

153KB Sizes 1 Downloads 62 Views

Engineering Failure Analysis 12 (2005) 654–661 www.elsevier.com/locate/engfailanal

Failure analysis of cold drawn prestressing steel wires subjected to stress corrosion cracking J. Toribio *, E. Ovejero Department of Materials Engineering, University of Salamanca, E.P.S., Campus Viriato, Avda. Requejo 33, 49022 Zamora, Spain Received 12 July 2004; accepted 26 December 2004 Available online 31 March 2005

Abstract This paper analyzes the stress corrosion performance of cold drawn prestressing steels with different degrees of cold drawing, in the two cases of pure stress corrosion cracking and fracture by hydrogen embrittlement. Results show a progressively anisotropic stress corrosion behaviour with cold drawing, and this anisotropy could be induced by the microstructural orientation. Consequences for engineering design are derived on the basis of the calculation of the transition stress intensity factor (from the stress corrosion subcritical crack growth to the critical cleavage propagation). While in situations of pure stress corrosion cracking such a value is even higher than the fracture toughness of the material in air (due to crack tip blunting), in the case of hydrogen embrittlement there is a marked reduction of stress intensity factor value, which indicates that these steels are highly susceptible to this phenomenon.  2005 Elsevier Ltd. All rights reserved. Keywords: Prestressing steel; Cold drawing; Stress corrosion cracking; Localised anodic dissolution; Hydrogen assisted cracking; Hydrogen embrittlement

1. Introduction Cold drawn prestressing steels, widely used in civil engineering as the main constituents of prestressed concrete structures, are manufactured from a previously hot rolled bar of pearlitic steel which is cold drawn in several passes to obtain the commercial prestressing steel wire with increased yield strength as a consequence of the activation of a strain-hardening mechanism. Macroscopically, cold drawing produces a progressive reduction in the diameter of the wire by both axial tensile stresses and transverse compressive stresses. Therefore, the final commercial product has undergone strong plastic deformations able to modify *

Corresponding author. Tel.: +34 980 54 50 00; fax: +34 980 54 50 02. E-mail address: [email protected] (J. Toribio).

1350-6307/$ - see front matter  2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.engfailanal.2004.12.021

J. Toribio, E. Ovejero / Engineering Failure Analysis 12 (2005) 654–661

655

drastically its microstructure and to induce progressive anisotropy in the material as a consequence of the important changes at the microsrructural level [1–6]. Thus, in spite of the fact that cold drawing enhances the classical mechanical properties of the steel – thereby providing a better material for structural engineering performance – the microstructural changes during manufacture may affect seriously the fracture behaviour in air [7–9] as well as the stress corrosion cracking (SCC) performance [10–15], which is potentially dangerous when the engineering structures work in aggressive environments. In this paper, the SCC performance of cold drawn prestressing steels with different degrees of cold drawing is analyzed in the two main regimes of environmentally assisted cracking: localised anodic dissolution and hydrogen assisted cracking, i.e., pure stress corrosion cracking (the former) and fracture by hydrogen embrittlement (the latter). In addition, since steels with different degrees of cold drawing are studied, the influence of the strain-hardening level on the stress corrosion behaviour can be elucidated.

2. Experimental programme The materials used in this work were high-strength steels taken from a real manufacturing process. Wires with different degrees of cold drawing were obtained by stopping the manufacturing chain and taking samples from the intermediate stages. The different steels were named with digits 0–6 which indicate the number of cold drawing steps undergone. Table 1 shows the chemical composition common to all steels, and Table 2 includes the diameter (Di), the cold drawing degree (represented by the ratio of the diameter of any steel to the initial diameter before cold drawing Di/D0), the yield strength (rY) the ultimate tensile stress rR and the fracture toughness (KIC) of the different steel wires. It should be noted here that, while the fracture behaviour of the slightly drawn steels (from 0 to 3) is isotropic, i.e., associated with mode I crack propagation, the most heavily drawn steels (from 4 to 6) exhibit anisotropic fracture behaviour with crack deflection and mixed mode propagation containing an important component in mode II [9]. Thus the KIC-value given in Table 2 represents only a measure of failure resistance in each steel wire, but it is an actual fracture toughness, i.e., a material property – only in the slightly drawn steels in which cracking develops in mode I, whereas in the case of the heavily drawn steels it plays the role of an ‘‘apparent’’ toughness which is useful for engineering design but not a material constant, since in this case mixed mode propagation appears, and the KIC-value was evaluated as if the crack propagation developed in mode I. The reason for this increasingly anisotropic fracture behaviour of the steels is the cold drawing process, which produces a microstructural orientation of the two basic microstructural units of the steels: the pearlitic colonies [4] and the pearlitic lamellae [6]. Both units tend to align parallel or quasi-parallel to the wire axis or cold drawing direction in the course of manufacture. Table 1 Chemical composition (wt%) of the steels C

Mn

Si

P

S

Cr

V

Al

0.80

0.69

0.23

0.012

0.009

0.265

0.060

0.004

Table 2 Diameter reduction and mechanical properties of the steels Steel

0

1

2

3

4

5

6

Di (mm) Di/D0 rY (GPa) rR (GPa) KIC (M Pam1/2)

12.00 1 0.686 1.175 60.1

10.80 0.90 1.100 1.294 61.2

9.75 0.81 1.157 1.347 70.0

8.90 0.74 1.212 1.509 74.4

8.15 0.68 1.239 1.521 110.1

7.50 0.62 1.271 1.526 106.5

7.00 0.58 1.506 1.762 107.9

656

J. Toribio, E. Ovejero / Engineering Failure Analysis 12 (2005) 654–661

To analyze the stress corrosion behaviour of the different steels, slow strain rate tests were performed on precracked steel wires. Samples were precracked by axial fatigue in the normal laboratory air environment to produce a transverse precrack, so that the maximum stress intensity factor during the last stage of fatigue precracking was Kmax = 0.30KIC, where KIC is the fracture toughness, and the crack depth was a = 0.30D in all cases, with D as the wire diameter. After precracking, samples were placed in a corrosion cell containing aqueous solution of l g/1 Ca(OH)2 plus 0.1 g/1 NaCl (pH 12.5) to reproduce the alkaline working conditions of prestressing steel surrounded by concrete. The experimental device consisted of a potentiostat and a three-electrode assembly (metallic sample or working electrode, platinum counter-electrode and saturated calomel electrode as the reference one). All tests were conducted under potentiostatic control, one group at a constant potential of 1200 mV vs. SCE at which the SCC mechanism is hydrogen assisted cracking (HAC) and another group at 600 mV vs. SCE at which the SCC mechanism is localised anodic dissolution (LAD), according to previous research on similar steels [12]. The applied displacement rate in the axial direction was constant during each test and proportional to each wire diameter so that the smallest rate was 1.7 · l03 mm/min for the fully drawn wire (steel 6) and the highest was 3.0 · l03 mm/min for the hot rolled bar (steel 0). The load applied on the sample was continuously monitored during the tests and the applied displacement was proportional to the time since the displacement rate was constant during the test.

3. Experimental results In Fig. 1 a plot is given of the time to final fracture in the tests, as well as the susceptibility of the different steels to HAC, evaluated as the ratio of the fracture load in the HAC test FHAC to the fracture load in air F0, as recommended by the ISO standard [16]. The results show no clear trend in the matter of susceptibility to HAC as a function of the degree of cold drawing (expressed by the ratio Di/D0), i.e., no significant improvement is detected in the fracture resistance of the steels as a consequence of the cold drawing process when the environmental mechanism is hydrogen embrittlement. However, a certain drawing-induced increase of time to failure seems to exist, so the steel performance is slightly better in the cold drawn prestressing steel wire (final product) than in the hot rolled bar (base material). Fig. 2 shows the evolution of the fracture profile in HAC as the degree of cold drawing increases. In the first steps of cold drawing (specimens 0 and 1) the crack growth develops in mode I in both fatigue precracking and HAC. In steels 2 and 3 there is a slight deflection in the hydrogen-assisted crack. For the most heavily drawn specimens (4–6) the crack deflection takes place suddenly after the fatigue precrack, and the

Fig. 1. Fracture load in HAC conditions (FHAC) divided by the fracture load in air (F0) and time to fracture in the tests (tHAC).

J. Toribio, E. Ovejero / Engineering Failure Analysis 12 (2005) 654–661

657

Fig. 2. Fracture profile of the different steels in the cathodic regime of cracking (HAC).

deviation angle is even higher. In these last stages of cold drawing, not only crack deflection but also crack branching is observed just after the fatigue precrack tip, i.e., there are two pre-damage directions (crack embryos), only one of which becomes the final fracture path. The mode I propagation distance associated with subcritical crack growth by HAC is given in Table 3. It only appears in the first stages of cold drawing, since in heavily drawn steels crack deflection takes place just at the fatigue precrack and thus the mode I propagation distance is zero. Fig. 3 gives the time to final fracture in the tests, as well as the susceptibility of the different steels to LAD, evaluated as the ratio of the fracture load in the LAD test FLAD to the fracture load in air F0. Although the scatter is high, the fracture load in the solution seems to increase with the degree of cold drawing (expressed by the ratio Di/D0). With regard to the time to failure the trend is even clearer, which indicates that a significant drawing induced improvement is detected in the fracture resistance of the steels when the environmental mechanism is anodic dissolution (or pure SCC), so the steel performance is remarkably better in the cold drawn prestressing steel wire than in the hot rolled bar. Fig. 4 shows the evolution of the fracture profile in LAD as the degree of cold drawing increases. As in the case of HAC, an increasingly anisotropic HAC behaviour with cold drawing is observed. For the slightly drawn steels (0–2), the fracture surfaces are macroscopically plane and oriented perpendicularly to the loading axis. Steel 3 shows a certain angle between the plane of the fatigue precrack and the fracture Table 3 Mode I propagation distance associated with subcritical crack growth by SCC (mm) Steel

0

1

2

3

4

5

6

HAC LAD

0.9 0.013

0.8 0.166

1.2 0.20

0.0 0.95

0.0 1.15

0.0 0.51

0.0 0.15

Fig. 3. Fracture load in LAD conditions (FLAD) divided by the fracture load in air (F0) and time to fracture in the tests (tLAD).

658

J. Toribio, E. Ovejero / Engineering Failure Analysis 12 (2005) 654–661

Fig. 4. Fracture profile of the different steels in the anodic regime of cracking (LAD).

propagation direction. In the most heavily drawn steels (4–6) the deviation from the fatigue precrack plane is even higher. The mode I propagation distance associated with subcritical crack growth by LAD is given in Table 3. It should be emphasised that in the case of LAD the steel is able to undergo mode I subcritical cracking even in the most heavily drawn steels, and this is another reason for the clear improvement of its corrosion resistance.

4. Fracture mechanics approach A fracture mechanics approach is presented in this section to evaluate a characteristic (or critical) stress intensity factor associated with the transition from the subcritical regime of SCC to the critical regime of environmentally unassisted fracture by cleavage. This transition stress intensity factor is a fundamental issue in engineering design against environmentally assisted fracture, To obtain it, an expression is required of the stress intensity factor KI for the geometry and loading mode under consideration: a cylinder in tension with an edge crack perpendicular to the tensile loading direction. The following expression [17] was used: pffiffiffiffiffiffi K I ¼ MðnÞr pa; ð1Þ where r is the remote axial stress (far from the crack), a the crack depth and M(n) a dimensionless function given by: 1=2

MðnÞ ¼ ð0:473  3:286n þ 14:797n2 Þ

1=4

ðn  n2 Þ

;

ð2Þ

where n is the ratio a/D of the crack depth to the sample diameter. This function comes from the computation – by the compliance method – of the global energy release rate in the considered geometry and loading mode. This approach is only valid to establish quantitative relationships in slightly drawn steels in which the fracture process develops in mode I in the subcritical and critical regimes. On the other hand, the crack deflection which takes place in heavily drawn steels produces a mixed mode stress state (with an important mode II component) so that the computation of stress intensity factors is an extremely difficult task, beyond the scope of this paper, since there is not only KI but also KII during the subcritical crack growth. However, results expressed in terms of the fracture load in aggressive environment, cf. Figs. 1 and 3, allow at least a roughly quantitative estimation – although not in the framework of fracture mechanics – of the susceptibility of heavily drawn steels to SCC. The transition stress intensity factor KT (in particular KHAC for HAC and KLAD for LAD) may be calculated as follows: pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ð3Þ K T ¼ Mðafat þ xSCC ÞrSCC pðafat þ xSCC Þ; where afat is the fatigue precrack (that existing at the beginning of the SCC test), xSCC the depth of subcritical crack growth by SCC in mode I, and rSCC the remote stress at the critical instant of the SCC tests (maximum values). Since xSCC was measured in direction perpendicular to the crack front, the critical crack depth is afat + xSCC. In addition, the critical remote stress rSCC is associated with the maximum load point in the load–displacement curve, i.e., with the instability point reached after the subcritical crack growth and

J. Toribio, E. Ovejero / Engineering Failure Analysis 12 (2005) 654–661

659

just before cleavage final fracture, which indicates that both the remote stress and the crack depth used to calculate KT in Eq. (3) correspond to the same physical situation: the critical instant in the SCC test. The distance xSCC is given in Table 3 for the two regimes of cracking, while the critical remote stress is: rSCC ¼

4F SCC ; pD2

ð4Þ

where FSCC is FHAC in Fig. 1 and FLAD in Fig. 3 in the respective cases of HAC and LAD.

5. Discussion The ratio KT/KIC (where KIC is the fracture toughness, cf. Table 2) allows a fracture mechanics evaluation of the SCC susceptibility of each steel. Fig. 5 represents the ratio KHAC/KIC as a function of the degree of cold drawing Di/D0 of slightly drawn steels, showing that these materials are very susceptible to hydrogen embrittlement, with a marked reduction of fracture resistance in these environmental conditions which could be enhanced, e.g., by cathodic overprotection [18–20]. This is a risk which should be avoided carefully in civil engineering design, construction and maintenance of prestressed concrete structures. In addition, Fig. 5 proves that the manufacturing process by cold drawing produces no significant improvement in the resistance of the steel to hydrogen embrittlement, in spite of the clear increase of yield strength that it imparts to the material, which is the final aim of this steelmaking technique. This seems to be valid in the case of heavily drawn steels, according to the results in Fig. 1. Furthermore, previous research work on fully drawn prestressing steel demonstrated that the very high yield strength could be responsible for the significant susceptibility of this material to hydrogen assisted cracking [21]. Fig. 6 shows the susceptibility KLAD/KIC as a function of the degree of cold drawing Di/D0 of slightly drawn steels. A remarkable increase of the SCC resistance of the steels with cold drawing is observed, so that the transition stress intensity factor is as high as KLAD = 1.61KIC in steel 2, i.e., clearly higher than the fracture toughness of the material in air. This tremendous increase of stress intensity level may be attributed to crack blunting in the near-tip due to the anodic dissolution process itself. As a matter of fact, the toughening effect of crack tip blunting has been reported previously in the matter of fracture in air [22,23] and SCC [24–26], and Fig. 6 indicates that the blunting effect increases with cold drawing, at least in slightly drawn steels.

Fig. 5. Susceptibility of slightly drawn steels to HAC, evaluated as the ratio of the critical stress intensity factor in the HAC test KHAC to the fracture toughness in air KIC.

660

J. Toribio, E. Ovejero / Engineering Failure Analysis 12 (2005) 654–661

Fig. 6. Susceptibility of slightly drawn steels to LAD, evaluated as the ratio of the critical stress intensity factor in the LAD test KLAD to the fracture toughness in air KIC.

In heavily drawn steels, although crack tip blunting should increase even more, this effect is interrupted by the anisotropic behaviour (cf. Fig. 4) which also increases with cold drawing. This could explain why the fracture load FLAD given in Fig. 3 decreases in the final drawing steps. Thus, the beneficial blunting effect is counterbalanced in heavily drawn steels by the mixed mode propagation following directions of minimum resistance to SCC. It is consistent with the results of Table 3 regarding the effect of cold drawing on the mode I subcritical crack growth distance which is increasing in slightly drawn steels and decreasing in heavily drawn steels.

6. Conclusions A fracture mechanics approach to stress corrosion cracking of cold drawn prestressing steels is presented in this paper, analyzing the particular phenomena of hydrogen assisted cracking (HAC) and localised anodic dissolution (LAD). A characteristic parameter for engineering design against environment-sensitive fracture is the critical stress intensity factor linked with the transition from the subcritical regime of environmental cracking to the critical regime of cleavage fracture. In the case of HAC there is a marked reduction of the stress intensity factor value, which indicates that cold drawn prestressing steels are very susceptible to hydrogen embrittlement. This dangerous process could be induced by cathodic overprotection. In LAD the transition stress intensity factor is even higher than the fracture toughness of the material in air, which indicates that these materials are highly resistant to stress corrosion cracking, probably as a consequence of a micromechanism of crack tip blunting enhanced by anodic dissolution in the vicinity of the crack tip.

Acknowledgements The authors thank the present financial support of their research at the University of Salamanca provided by the following institutions: Spanish Ministry for Scientific and Technological Research MCYTFEDER (Grant No. MAT2002-01831), FEDER-INTERREG III (Grant No. RTCT-B-Z/SP2.P18), Junta

J. Toribio, E. Ovejero / Engineering Failure Analysis 12 (2005) 654–661

661

de Castilla y Leo´n (JCYL; Grant No. SA078/04) and Spanish Foundation ‘‘Memoria de D. Samuel Solo´rzano Barruso’’.

References [1] Embury JD, Fisher RM. The structure and properties of drawn pearlite. Acta Metall 1966;14:147–59. [2] Langford G. A study of the deformation of patented steel wire. Metall Trans 1970;1:465–77. [3] Toribio J, Ovejero E. Microstructure evolution in a pearlitic steel subjected to progressive plastic deformation. Mater Sci Eng 1997;A234-236:579–82. [4] Toribio J, Ovejero E. Microstructure orientation in a pearlitic steel subjected to progressive plastic deformation. J Mater Sci Lett 1998;17:1037–40. [5] Toribio J, Ovejero E. Effect of cumulative cold drawing on the pearlite interlamellar spacing in eutectoid steel. Scripta Mater 1998;39:323–8. [6] Toribio J, Ovejero E. Effect of cold drawing on microstructure and corrosion performance of high-strength steel. Mech TimeDependent Mat 1998;1:307–19. [7] Astiz MA, Valiente A, Elices M, Bui HD. Anisotropic fracture behaviour of prestressing steels. In: Faria LO, editor. Life assessment of dynamically loaded materials and structures-ECF5. West Midlands: EMAS; 1984. p. 385–93. [8] Toribio J, Toledano M. Fatigue and fracture performance of cold drawn wires for prestressing concrete. Constr Build Mater 2000;14:47–53. [9] Toribio J, Ovejero E, Toledano M. Microstructural bases of anisotropic fracture behaviour of heavily drawn steel. Int J Fracture 1997;87:L83–8. [10] Cherry BW, Price SM. Pitting, crevice and stress corrosion cracking studies of cold drawn eutectoid steels. Corros Sci 1980;20:1163–84. [11] Langstaff DC, Meyrick G, Hirth JP. Hydrogen induced delayed failure of high strength alloy steel wires. Corrosion-NACE 1981;37:429–37. [12] Parkins RN, Elices M, Sa¨nchez-Ga¨lvez V, Caballero L. Environment sensitive cracking of pre-stressing steels. Corros Sci 1982;22:379–405. [13] Sarafianos N. Environmentally assisted stress-corrosion cracking of high-strength carbon steel patented wire. J Mater Sci Lett 1989;8:1486–8. [14] Price SM, Warden PG, Chia TS, Scarlett NVY, Mahoney JH. The role of notches in the hydrogen-assisted cracking of steel prestressing tendons. Mater Forum 1992;16:245–51. [15] Toribio J, Lancha AM. Effect of cold drawing on environmentally assisted cracking of cold drawn steel. J Mater Sci 1996;31:6015–24. [16] ISO 7539-7. Corrosion of metals and alloys – stress corrosion testing. Part 7: slow strain rate testing. Geneva, Switzerland: ISO; 1989. [17] Valiente A. Criterios de fractura para alambres. PhD Thesis, Polytechnic University of Madrid, Spain; 1980. [18] Isecke B, Mietz J. The risk of hydrogen embrittlement in high-strength prestressing steels under cathodic protection. Steel Res 1993;64:97–101. [19] Congleton J, Hussain I, Parkins RN. Effect of applied potential on corrosion fatigue of wire ropes in sea water. Br Corros J 1985;20:5–18. [20] Hamano R. Effects of cathodic overpotential on fatigue crack growth behaviour of high strength steels in a corrosive environment. J Mater Sci Lett 1989;8:852–4. [21] Toribio J, Lancha AM. Effect of cold drawing on susceptibility to hydrogen embrittlement of prestressing steel. Mater Struct 1993;26:30–7. [22] Ritchie RO, Horn RM. Further considerations on the inconsistency in toughness evaluation of AISI 4340 steel austenitized at increasing temperatures. Metall Trans 1978;9A:331–41. [23] Handerhan KJ, Garrison Jr WM. A study of crack tip blunting and the influence of blunting behavior on the fracture toughness of ultra high strength steels. Acta Metall Mater 1992;40:1337–55. [24] Creager M, Paris PC. Elastic field equations for blunt cracks with reference to stress corrosion cracking. Int J Fracture Mech 1967;3:247–52. [25] Chu WY, Hsiao CM, Li SQ. A new engineering fracture toughness parameter KISCC(q). Scripta Metall 1979;13:1057–62. [26] Ray KK, Rao GR. A new test principle for determining threshold stress intensity factor KIEAC in environmentally assisted cracking. Int J Fracture 1993;61:R69–75.