1

AE MONITORING AND THREE DIMENSIONAL OBSERVATION OF

STRESS CORROSION CRACKING FROM CORROSION PITS

Kaige Wu1, Kaita Ito2, Ippei Shinozaki3, Manabu Enoki1

1 Department of Materials Engineering, The University of Tokyo, 7-3-1 Hongo,

Bunkyo-ku, Tokyo 113-8656, Japan 2 National Institute for Materials Science, Tsukuba 305-0047, Japan

3 Research Laboratory, IHI Corporation, Yokohama 235-8501, Japan

Phone: +81 3 5841 7129; fax: +81 3 5841 7181;

E-mail: kaigewu@rme.mm.t.u-tokyo.ac.jp

Abstract:

Stress corrosion cracking (SCC) is a complicated three-dimensional (3D) process involving both electrochemical and mechanical phenomena. The source mechanism of acoustic emission (AE) behavior during SCC development is still not well understood. In this work, AE monitoring combined with 3D characterization using X-ray computed tomography (X-ray CT) were proposed to investigate the SCC of an initially smooth sample of SUS420J2 stainless steel exposed to both a 0.6% constant strain and a neutral sodium chloride droplet at room temperature. Cracks were observed to originate from corrosion pits. Further 3D observation of the inner geometry of cracks indicated that cracks preferentially attacked the mouth, rather than the bottom, of the corrosion pit. It was also observed that by the path of crack propagation existed several interior corrosion pits which may play a critical role in crack growth. AE signals were detected during SCC evolution and characteristics of AE events were associated with crack initiation and propagation. K-means clustering analysis grouped the AE events into three clusters. With in-situ surface observation and 3D images of cracks, hydrogen bubble evolution, plastic deformation around the crack tip, and the crack propagation were considered as the possible AE source mechanisms during the SCC.

Keywords: stress corrosion cracking; corrosion pit; clustering; acoustic emission; X-ray CT

1. Introduction

Structural materials, especially stainless steels, are critical for developing and maintaining infrastructures. However, it can undergo stress corrosion cracking (SCC) by slow, environmentally induced crack initiation and propagation. The SCC process is concurrent by both electrochemical and mechanical phenomena that are the results of the synergistic effect of stress and corrosive environment on materials [1]. In many applications, corrosion pit serves the primary precursor to crack initiation and the general stages of SCC are therefore recognized as pitting corrosion, pit-to-crack transition and subsequent crack growth [2]. Regarding the crack initiation from corrosion pits, there still exist different arguments. Conventional viewpoint argues that SCC preferentially initiates at the base of a pit because the pit base not only serves the local stress-strain concentrator but also retains an aggressive environment [2-3]. Comparatively, another new perspective, based on the three dimensional (3D) observations with X-ray computed tomography (X-ray CT, or XCT), is put forward recently that cracks are not necessarily initiated at the pit base during the pit-to-crack

Mor

e in

fo a

bout

this

art

icle

: ht

tp://

ww

w.n

dt.n

et/?

id=

2357

9

2

transition [4]. It suggests that cracks nucleate on the pit walls in relation to the dynamic strain provided by the growing pit. Subsequently, cracks grow around the pit and finally coalesce to form a complete through-crack. The inconsistency implies further research work is needed on the pit-to-crack transition during the early stage of SCC.

On the other hand, developing in-situ methodologies is important in understanding the cracking mechanism and evaluation of the health condition of metallic structure and components. Acoustic emission (AE) is an in-situ nondestructive evaluation (NDE) method that bases on the detection of elastic waves generated by the fast energy release during an irreversible change within materials and solids. AE method holds great potential in monitoring the progression of materials degradation and steel corrosion [5-10]. AE method has long been applied to study the crack initiation and propagation during SCC. In the literature, different damage mechanisms during SCC, including the rupture of pit cover [10], fracture of metallic ligaments [11-12], plastic deformation around the crack tip [11, 13-18], ductile fracture [14], hydrogen bubble evolution [10, 16-21], falling off of the surface grain [20-21], rupture of corrosion products [20-22], and the cracking process [11-24] have been correlated with AE signals. However, the interpretation of the source mechanisms in current reports was mainly based on speculation of AE evolution coupled with post mortem microstructural analysis. Therefore, an irrefutable elucidation of the source mechanisms of the detected AE signals requires more investigation of in-situ visualized linkage. Moreover, the SCC testing proposed in literature [13-14, 18] is difficult to directly correlate the detected AE signals to one single step of cracking evolution since multiple-pits and cracks may be initiated to develop at different steps when the samples were immersed in the corrosive solution.

In order to better understand the crack initiation, we developed an experimental methodology to study the evolution of one single crack under a NaCl solution micro-droplet. Simultaneously, AE monitoring accommodated with an in-situ surface observation was proposed to perform with the SCC testing. XCT was also used to give a 3D morphological observation of the crack and corrosion pit.

2. Experimental procedures

2.1 SCC testing and AE measurements

The material used here is SUS420J2 (C: 0.32 mass%, Si: 0.45 mass%, Mn: 0.54 mass%, Cr: 13.39 mass%, Ni:0.09 mass%, P: 0.031 mass%, S: 0.005 mass%, Fe : bal.). Three kinds of samples were prepared with different heat treatment sequences as follows. (Sample A): 600°C/1h → 800°C/1h → 1010°C/2.75h → gas cooling → 180°C/3h; (Sample B): 600°C/1h → 800°C/1h → 1010°C/2h45m →gas cooling; (Sample C): 950°C/2h → gas cooling. The Vicker’s hardness of sample A, B, and C were HV 593, 611, and 511. Then they were polished to a mirror-like finish. The SCC samples were fixed in a jig with a curvature of R125 to apply a constant strain of 0.6% to the sample surface. A micro-droplet of 1.0 mass% NaCl solution of 1μL was set in the center of the surface of samples placed inside a thermostatic bath (298K, 99-100%). The bath was capped with a transparent glass cover to accommodate an in-situ surface observation with a digital microscope (VHX-5000, Keyence Corp., Japan). Two high-sensitivity R-CAST sensor systems of 200 kHz resonant (M204A, Fuji Ceramics, Japan) were respectively placed 10 mm and 30 mm away from the droplet site to filter out the noise outside the corrosion region by source location method. A continuous acquisition mode named “Continuous Wave Memory” (CWM) which was developed by our group [25] was used to record and analyze the AE signals. The sampling rate and threshold were 5

3

MHz and 20 mV, respectively. The experimental arrangements are schematically shown in Fig. 1. Finally, the extracted AE data were analyzed by k-means clustering algorithm [26].

2.2 3D observation of crack morphology

After testing, the SCC samples were both inspected by SEM and XCT. The SCC samples were polished and cut to a small size of XCT sample, schematically shown in Fig. 2(a), to meet the requirement of lab XCT scanning. A 3D XCT Scanner (TDM1000-IS, Yamato Scientific Co., Ltd., Japan) was used for scanning the internal structural information of cracks. The X-ray computed tomographic process is schematically shown in Fig. 2(b). The XCT sample was fixed on a rotational stage. A rotation step of 0.24 degree and 25 mins per full rotation was set for high-resolution scanning. The X-ray tube voltage was set to 100 kV and

Fig. 1 Schematic of the AE measurements and SCC test incorporated with an in-situ digital microscope system.

(a)

(b)

Fig. 2 (a) Schematic of the preparation of XCT sample, (b) 3D X-ray computed tomographic process of internal view of cracks.

4

X-ray tube current was set to 30μA~54μA in dependence of crack size. 2D-slice images were reconstructed using the image processing unit of the X-ray CT system. Finally, the volumetric image visualization and segmentation of the crack morphology were implemented using VoTracer software package (Riken, Japan).

3. Results

3.1 Surface and 3D observation of cracks

Fig. 3 shows the surface morphologies of cracks in three samples. First, one single crack was initiated under the micro-droplet. Regardless of the heat-treatment exposures, the cracks of all samples were observed to stem from the corrosion pits. In detail, the cracks developed preferentially at the place next to the pit mouth on the sample surface, rather than the general perspective that cracks develop preferentially and coalesce at the pit base [2-3].

Fig. 4 Reconstructed tomographic 2D slices and 3D volume rendering iamges of cracks morphologies of sample A, B, and C.

Fig. 3 SEM observations of the crack morphologies of samples (a) A, (b) B, and (c) C.

5

Fig. 4 shows the reconstructed slices and volume rendering images. The internal views provide a direct illustration of the relative position of the crack and corrosion pit, i.e., the crack is next to but does not cross through the corrosion pit. This observation is close to the recently-proposed viewpoint that the cracks primarily attack the mouth, rather than the bottom, of the corrosion pits [4]. But the different point is that the crack does not finally cross or combine the pit, unlike the formation of through-crack after crack growth rate exceeds the pit growth rate [2-4]. This may elucidate a different role of corrosion pits in the crack initiation in high strength steels. The volume rendering images indicate that the crack morphologies are not continuous in spatial distribution. The discreteness may imply a competition among different advancing crack tips which probably depended on the local inhomogeneity of the heat treated material. Furthermore, several interior cavities (named here “interior pit”) were observed to situate by the path of the crack propagation. One is shown in Fig. 5 by the reconstructed tomographic 2D slices. The interior pit is apparently an included local that may play a critical role in the crack growth.

3.2 AE signals and damage identification

Fig. 6 shows the evolution of AE signals obtained during SCC process of three samples. According to the increase trend and amplitude distribution, the data could be divided into four stages. Combined with the in-situ observational SCC evolution, the four-stage AE signals are associated with pitting corrosion, pit-to-crack transition, small crack growth and long crack propagation. Such observation is consistent with the fundamental steps of overall SCC development [1]. In stage I, chloride ions locally destroy the passive film surface of stainless steel sample, exposing the bare metal to the electrolyte and allowing the activation of localized corrosion. Later, local acidification by the hydrolysis of dissolving metal cations within pit lowers the pH value and generates the hydrogen bubbling which was captured in this work. A sharp increase in AE events in stage III reflects the transition from pit to crack. Stage II is a quiescence phase for sample A and B, but sharp increase for sample C. This stage seems to be microstructure dependent and will be further investigated in future work. Stage IV with the relatively low number of AE events is corresponding with crack propagation.

Fig. 5 Reconstructed tomographic slices showing an “interior pit” along the crack path of the sample C. The inserted schematic indicates the planes.

6

4000 6000 8000 10000 12000 14000 1600040

50

60

70

80

90

100(Sample-A)

AE amplitude

Cu

mu

lati

ve

AE

ev

ents

, N

AE

Am

pli

tud

e, d

B

Time, t / s

100

101

102

103

104

AE events

II

Stage I: Pit corrosion.

Stage II: Pit-to-crack transition.

Stage III: Small crack growth.

Stage IV: Crack propagation.

IVIII

Stage I

46000 48000 50000 52000 5400040

50

60

70

80

90

100 AE amplitude

Time, t / s

Cu

mu

lati

ve

AE

ev

ents

, N

100

101

102

103

104

II

(Sample-B)

AE events

AE

Am

pli

tud

e, d

B

Stage I

IVIII

50000 100000 150000 200000 25000040

50

60

70

80

90

Cu

mu

lati

ve

AE

ev

ents

, N

AE

am

pli

tud

e, d

B

Time, t / s

(Sample-C)

AE amplitudeIII IVII

Stage I

100

101

102

103

104

Cumulative AE events

Fig. 6 Evolution of cumulative events and amplitude of AE signals obtained during SCC evolution of samples A, B, and C.

Fig. 7 Identification of AE source mechanisms during SCC from in-situ surface observation of the crack evolution (sample A).

7

To identify the damage mechanism for AE signals, k-means clustering algorithm was used to analyze the AE data. According to the similarity of AE data, three different AE clusters were obtained. From in-situ linkage analysis, three AE clusters were associated with hydrogen bubble evolution, plastic deformation and the cracking process. Fig. 7 shows the partial data of distribution of three AE clusters and the corresponding observation of a periodic coalescence phenomenon of two crack branches during crack propagation. Interestingly, high-amplitude signals were highly correlated with the occurrence of coalescence events of two branches. Therefore, the cluster 3 (red) is further attributed to the crack-branches coalescence. Suggestively, the plastic deformation and hydrogen bubble evolution dominate the detected AE signals during the crack growth. The cracking itself seems to be “silent” until the coalescence events occur with the concurrent high-amplitude signals. Moreover, the high-amplitude signals are not continuous in time series. This may be a response to the discrete morphologies of cracks shown in Fig. 4.

4. Discussion

The direct optic capture of the hydrogen bubble evolution approaching the point of crack initiation suggests that the hydrogen embrittlement mechanism may dominate the crack initiation of the high-strength steels in the present work. First, it is well known that any process producing hydrogen at a metal surface will induce considerable atomic hydrogen absorption and rapid diffusion into the metal lattice even at normal temperature [27]. The hydrogen diffusion phenomenon is more preferable in high-strength steels and enhanced by the applied tensile stress. The observed surface- and interior-pits can serve the source of hydrogen through the “local acidity” by hydrolysis of the dissolved metal ions [28]. Also, a previous report explained the SCC with hydrogen embrittlement mechanism accompanied by AEs in 13% Cr martensitic stainless steel exposed to 3% NaCl solution [29]. Furthermore, considering the observed zigzag-feature, the crack presents the intergranular type. Hydrogen is but one element that can segregate to grain- or interphase boundaries and its solubility in a metal depends on the applied tensile stress level [30]. This may account for that the cracks primarily attacked the pit/grain boundary interface on the sample surface, rather than the pit bottom. As the surface of bent sample suffered the highest stress level, which may cause the site near surface being the preferential crack nuclei. Furthermore, the final cracks did not cross through or combine with the corrosion pits, suggesting that the critical role of corrosion pits may be the source of hydrogen in the pit-to-crack transition in high-strength steels.

On the other hand, the distribution of high-amplitude cracking events with the plastic deformation events can be explained by the Hydrogen-Enhanced Localized Plasticity (HELP) mechanism [31]. The cracking was followed by the generation and accumulation of dislocation by the combination of hydrogen atoms and stress. Therefore, the plastic deformation dominated the AE signals of medium amplitude until the occurrence of crack-branches coalescence events which produced the strongest AE signals. This is consistent with the conclusions of previous reports [16-17] that the crack extending along the grain boundaries cannot generate detectable AE signals. Further work is required to give a reliable validation of this explanation.

5. Conclusion

The behavior of one single stress corrosion crack of 13Cr martensitic stainless steel was controlled and investigated by AE monitoring and 3D XCT observation. The crack was

8

observed to primarily originate from the mouth of a corrosion pit. The hydrogen embrittlement is supposed to account for the mechanism of crack initiation of SUS420J2. The corrosion pit may serve a critical role of providing the hydrogen in pit-to-crack transition. The evolution of AE signals were stage-by-stage associated with the crack initiation and propagation. By direct linkage with in-situ surface observations, the AE signals were attributed to the hydrogen bubble evolution, plastic deformation, and the coalescence events of crack-branches. However, a valid interpretation of the AE signals requires in-situ 3D visualizations of the crack evolution and quantitative evaluations of the crack behavior with AE data.

Acknowledgements

This work was supported by Council for Science, Technology and Innovation (CSTI), Cross-ministerial Strategic Innovation Promotion Program (SIP), “Structural Materials for Innovation” (Funding agency: JST). The authors acknowledge the great helps from Prof. Jun Takahashi, Dr. Mizuki Ohta, Dr. Pornthep Chivavibul, Dr. Takayuki Shiraiwa in X-ray CT scanning and image processing.

References:

[1] Parkins, R.N., Stress corrosion spectrum, British Corrosion Journal, Vol. 7, 1972, pp. 15-

28.

[2] Turnbull, A., Zhou, S., Orkney, L., McCormick, N., Visualization of stress corrosion cracks

emerging from pits, Corrosion, Vol. 62, 2006, pp. 555-558.

[3] Parkins, R.N., Singh, P.M., Stress corrosion crack coalescence, Corrosion, Vol. 46, 1990,

pp. 485-499.

[4] Horner, D.A., Connolly, B.J., Zhou, S., Crocker, L., Turnbull, A., Novel images of the

evolution of stress corrosion cracks from corrosion pits, Corrosion Science, Vol. 53, 2011, pp.

3466-3485.

[5] Ono, K., Acoustic Emission in Materials Research – A Review, Journal of Acoustic

Emission, Vol. 29, 2011, pp. 284-308.

[6] Huang, M., Jiang, L., Liaw, P.K., Brooks, C.R., Seeley, R., Klarstrom, D.L., Using acoustic

emission in fatigue and fracture materials research., JOM, Vol. 50, 1998, pp. 1-14.

[7] Wu, K.G., Choe, C.Y., Lee, S.M., Yoo, J., Hyun, C.Y., Hong, D.P., Byeon, J.W., Real-time

Monitoring of Degradation of Zirconia Ceramic by Acoustic Emission, Applied Mechanics and

Materials, Vol. 433, 2013, pp. 816-820.

[8] Wu, K., Jung, W.S., Byeon, J.W., Evaluation of corrosion critical variables of 304 stainless

steel by delay time of acoustic emission, Materials Transactions, Vol. 56, 2015, pp. 398-403.

[9] Wu, K., Jung, W.S., Byeon, J.W., Acoustic emission of hydrogen bubbles on the counter

electrode during pitting corrosion of 304 stainless steel, Materials Transactions, Vol. 56, 2015,

pp. 587-592.

9

[10] Wu, K., Jung, W.S., Byeon, J.W., In-situ monitoring of pitting corrosion on vertically

positioned 304 stainless steel by analyzing acoustic-emission energy parameter, Corrosion

Science, Vol. 105, 2016, pp. 8-16.

[11] Sung, K.Y., Kim, I.S., Yoon, Y.K., Characteristics of acoustic emission during stress

corrosion cracking of inconel 600 alloy, Scripta materialia, Vol. 37, 1997, pp. 1255-1262.

[12] Alvarez, M.G., Lapitz, P., Ruzzante, J., AE response of type 304 stainless steel during

stress corrosion crack propagation, Corrosion Science, Vol. 50, 2008, pp. 3382-3388.

[13] Kovač, J., Leban, M., Legat, A., Detection of SCC on prestressing steel wire by the

simultaneous use of electrochemical noise and acoustic emission measurements,

Electrochimica Acta, Vol. 52, 2007, pp. 7607-7616.

[14] Kovac, J., Alaux, C., Marrow, T.J., Govekar, E., Legat, A., Correlations of

electrochemical noise, acoustic emission and complementary monitoring techniques during

intergranular stress-corrosion cracking of austenitic stainless steel, Corrosion Science, Vol.

52, 2010, pp. 2015-2025.

[15] Shaikh, H., Amirthalingam, R., Anita, T., Sivaibharasi, N., Jaykumar, T., Manohar, P.,

Khatak, H.S., Evaluation of stress corrosion cracking phenomenon in an AISI type 316LN

stainless steel using acoustic emission technique, Corrosion Science, Vol. 49, 2007, pp. 740-

765.

[16] Xu, J., Han, E.H., Wu, X., Acoustic emission response of 304 stainless steel during

constant load test in high temperature aqueous environment, Corrosion Science, Vol. 63,

2012, pp. 91-99.

[17] Xu, J., Wu, X., Han, E.H., Acoustic emission response of sensitized 304 stainless steel

during intergranular corrosion and stress corrosion cracking, Corrosion Science, Vol. 73,

2013, pp. 262-273.

[18] Calabrese, L., Bonaccorsi, L., Galeano, M., Proverbio, E., Di Pietro, D., Cappuccini, F.,

Identification of damage evolution during SCC on 17-4 PH stainless steel by combining

electrochemical noise and acoustic emission techniques, Corrosion Science, Vol. 98, 2015,

pp. 573-584.

[19] Du, G., Li, J., Wang, W.K., Jiang, C., Song, S.Z., Detection and characterization of

stress-corrosion cracking on 304 stainless steel by electrochemical noise and acoustic

emission techniques, Corrosion Science, Vol. 53, 2011, pp. 2918-2926.

[20] Yonezu, A., Cho, H., Ogawa, T., Takemoto, M., Simultaneous monitoring of acoustic

emission and corrosion potential fluctuation for mechanistic study of chloride stress corrosion

cracking, Key Engineering Materials, Vol. 321, 2006, pp. 254-259.

[21] Yonezu, A., Cho, H., Takemoto, M., Monitoring of stress corrosion cracking in stainless

steel weldments by acoustic and electrochemical measurements, Measurement Science and

Technology, Vol. 17, 2006, pp. 2447-2454.

10

[22] Shiwa, M., Masuda, H., Yamawaki, H., Ito, K., Enoki, M., In-Situ Observation and

Acoustic Emission Analysis for SCC of MgCl2 Droplet in SUS304 Stainless Steel, Materials

Transactions, Vol. 55, 2014, pp. 285-289.

[23] Shiwa, M., Yamawaki, H., Masuda, H., Ito, K., Enoki, M., AE signals analysis during

chloride droplet SCC on thin plate of SUS304 steel, Strength, Fracture and Complexity, Vol.

5, 2009, pp. 109-116.

[24] Ito, K., Yamawaki, H., Masuda, H., Shiwa, M., Enoki, M., SCC monitoring of chloride

droplets on thin SUS304 plate specimens by analysis of continuous recorded AE waveform,

Materials transactions, Vol. 51, 2010, pp. 1409-1413.

[25] Ito, K., Enoki, M., Acquisition and analysis of continuous acoustic emission waveform for

classification of damage sources in ceramic fiber mat, Materials transactions, Vol. 48, 2007,

pp. 1221-1226.

[26] Muto, Y., Shiraiwa, T., Enoki, M., Evaluation of the deformation behavior in directionally

solidified Mg–Y–Zn alloys containing LPSO phases by AE analysis, Materials Science and

Engineering: A, Vol. 689, 2017, pp. 157-165.

[27] Rogers, H.C., Hydrogen Embrittlement of Metals: Atomic hydrogen from a variety of

sources reduces the ductility of many metals, Science, Vol. 159, 1968, pp. 1057-1064.

[28] Ryan, M.P., Williams, D.E., Chater, R.J., Hutton, B.M., McPhail, D.S., Why stainless

steel corrodes, Nature, Vol. 415, 2002, pp. 770-774.

[29] Okada, H., Yukawa, K.I., Tamura, H., Application of acoustic emission technique to the

study of stress corrosion cracking in distinguishing between active path corrosion and

hydrogen embrittlement, Corrosion, Vol. 30, 1974, pp. 253-255.

[30] Courtney, T.H., Embrittlement, in: Courtney, T.H. (Eds.), Mechanical Behavior of

Materials, Second ed., McGraw-Hill International Editions, Singapore, 2000, pp. 630-685.

[31] Robertson, I.M., Sofronis, P., Nagao, A., Martin, M.L., Wang, S., Gross, D.W., Nygren,

K.E., Hydrogen embrittlement understood, Metallurgical and Materials Transactions A, Vol.

46, 2015, pp. 2323-2341.