Accepted Manuscript

Corrosion Fatigue Crack Propagation in a Heat Affected Zone of High-Per‐

formance Steel in an Underwater Sea Environment

Dong-Hwan Kang, Jong-Kwan Lee, Tae-Won Kim

PII: S1350-6307(10)00149-4
DOI: 10.1016/j.engfailanal.2010.08.019
Reference: EFA 1432

To appear in: Engineering Failure Analysis

Received Date: 30 July 2010

Accepted Date: 25 August 2010

Please cite this article as: Kang, D-H., Lee, J-K., Kim, T-W., Corrosion Fatigue Crack Propagation in a Heat Affected
Zone of High-Performance Steel in an Underwater Sea Environment, Engineering Failure Analysis (2010), doi:
10.1016/j.engfailanal.2010.08.019

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
 Corrosion Fatigue Crack Propagation in a Heat Affected Zone of
High-Performance Steel in an Underwater Sea Environment
Dong-Hwan Kanga, Jong-Kwan Leeb and Tae-Won Kima *
a
School of Mechanical Engineering, Hanyang University, 17 Haengdang-dong,
Seongdong-gu, Seoul 133-791, Republic of Korea
b
Steel Structure Research Laboratory, Research Institute of Industrial Science & Technology,
Hwaseong 445-813, Republic of Korea

*
Tel.: +82 2 2220 0421; fax: +82 2 2292 0401; E-mail: twkim@hanyang.ac.kr

Abstract

Fatigue crack propagation in high-performance steel, HSB800, was investigated in air and

seawater environments using three-point single-edge notched bending fatigue tests. Two types of

heat affected zones (HAZ) resulting from welding processes, coarse grained and inter-critically

reheated coarse grained zones, together with as-received metal were used to analyze the behavior. A

3.5% sodium chloride solution was used as the seawater, and several types of loading conditions,

according to the stress ratio and loading frequency, were applied. The fatigue crack propagations

were then assessed using the da/dN- K curves. The results showed that the corrosion fatigue crack

in seawater was more rapidly propagated than was that in the air environment. The third region of

crack propagation was not observed in the da/dN- K curve for the seawater environment,

particularly in base metal. Generally, high stress ratio and load frequency shows a low threshold

value in the stress intensity factor range. Furthermore, a different microstructural aspect of the

crack propagation behavior in HAZs, as compared with that of the base metal, has been identified

according to the crack propagation mechanisms.

Keywords: Corrosion fatigue crack propagation; Heat affected zone; Fracture mechanics; Seawater;

1
High performance steel

1. Introduction

Recently, increasing demand for mega-structural applications, such as long-span bridges,

skyscrapers, and huge vessels has lead to the development of high-performance steels.

Development of high-performance steel requires much attention to high strength, high ductility,

good weldability and weatherability due to the severe environmental requirements. In particular,

securing of fracture toughness, together with weather resistance of the materials, to suppress fatigue

crack propagation (FCP) related to corrosion damage in complex environments is one of the most

important issues in the development of high-performance steel.

Steel architectural structures are typically composed of a large number of welding sub-

structures that are constructed via high temperature processes. As a result, the structural materials

may contain heat affected zones (HAZ) which usually have reduced mechanical properties due to

the modified grains. Furthermore, huge structures, such as long-span bridges, are often constructed

seaside because of geopolitical considerations in the design process. Thus, the materials used in

such structures are required to possess a substantial amount of corrosion resistance.

In general, the fatigue and corrosion fatigue resistances of a structural material are

influenced by various loading factors, such as load ratio and frequency. Regarding the load ratio,

Schmidt et al. [1] and Murakami et al. [2] have reported that the fatigue crack propagation rate

(FCPR) decreases with increasing stress ratio in high-strength steels. Different results, however,

were presented by Kujawski [3] and Sadananda et al. [4], who reported that the FCPR increases

with increasing stress ratio, particularly for aluminum alloys. In contrary to all of these views, Liaw

et al. [5] and King [6] have demonstrated that the load ratio has very little or no influence on the

2
FCPR behavior of Ni-based superalloys and stainless steels in air environments. Meanwhile, the

effect of load frequency on the FCPR in structural materials was discussed as follows. In general,

lower frequency has been considered to increase the corrosion fatigue crack propagation rate

(CFCPR) up to a certain frequency limit [7]. As much previous research has demonstrated, the

effect of such factors on the FCP or corrosion fatigue crack propagation (CFCP) behavior is largely

material and environment-dependent. Thus, proper characterizations of the fatigue and corrosion

fatigue behaviors of structural materials, particularly those used in severe environments are very

important [8-11].

In this work, experimental investigations of the FCP and the CFCP behaviors for the newly

developed high-performance steel, HSB800, were conducted in air and 3.5% NaCl solution to

represent a seawater environment. Two types of HAZ materials, together with the base metal, were

tested under various load ratio and frequency. Comparisons of the FCP and the CFCP behaviors of

each material were demonstrated, and fracto-graphs were constructed using optical microscopic and

scanning electron microscopic (SEM) observations to analyze the microstructural aspects of the

fracture process. A different microstructural aspect of the crack propagation behavior in HAZs, as

compared with that of the base metal, has been identified according to the crack propagation

mechanisms.

2. Experimental

High-performance steel for bridges, HSB800 (KS D 3868:2009), produced using a thermo-

mechanical control process (TMCP), was used as the base material. The thicknesses of the as-rolled

steel plates used in this work were 22 mm. The chemical composition and mechanical properties in

the rolling direction of the material are summarized in Tables 1 and 2, respectively. Mean grain

3
sizes of the base metal and the coarse-grained HAZ were approximately 3.5 m and 20 m,

respectively.

Single-edge notched bend (SENB) specimens with cross-sections of 9 mm x 15 mm were

extracted from the rolled-plates, with the notches in the L-T orientation. In order to analyze the

effects of HAZ on the CFCP behavior, two types of HAZ material, such as coarse-grained HAZ

(CG-HAZ) and inter-critically reheated coarse grained HAZ (ICCG-HAZ), were developed. A

Gleeble-1500 machine was used to produce the two types of HAZ through the application of 30

KJ/cm heat. The HAZs were located at the center of the notch within a 4 mm range in the middle of

specimen. After preparation of the base metal and the HAZ specimens, pre-fatigue cracks were

produced at a lower level of stress in air than that used in the actual test. This process was

controlled according to the standard, 'ISO-11782-2: Corrosion of metals and alloys [Corrosion

fatigue testing] Part 2 Crack propagation testing using pre-cracked specimens.'

The FCPR tests were conducted in air and seawater, in which the range of the stress

intensity factor, the ∆K envelope, was increased in a universal 50 KN mechanical testing

machine at room temperature. The machine was equipped with a digital controller interfaced to a

personal computer. A specially fabricated acryl container with an inverted three-point bend fixture

jig was set up to prevent the undesirable chemical reaction between the NaCl solution and the jig.

The loading jig and three rollers were composed of stainless steel and Teflon, respectively. A

schematic diagram for the testing apparatus is shown in Figure 1.

Using a commercialized crack gauge, KV-5C (KYOWA ELECTRONIC INSTRUMENTS

Co., LTD), instead of a crack opening displacement (COD) gauge, whose durability and

performance in aqueous environments has yet to be validated, the evolution of crack length was

directly measured. The crack gauge, attached to the surface of the specimen, was used to determine

the evolution of crack length by measuring the change in electric resistance corresponding to the

4
consecutive amputation of existing copper wires. All crack gauges used in the seawater

environment were covered with waterproof rubber tape.

Load frequency f and load ratio R were varied from 0.1 to 0.5 by employing triangular

waveforms. Based on the principle design of experiment (DOE), a carefully organized loading

condition was established according to the different types of materials, including the base metal,

CG-HAZ and ICCG-HAZ, to minimize the number of test cases. A summary of the combinations

of load ratio and frequency are shown in Table 3.

3. Results and discussions

Figures 2(a), 2(b) and 2(c) show the initial specimens, and the final specimens in air and

seawater, respectively. Figures 2(b) and 2(c) show that the final specimens did not obviously divide

into two parts in either air or seawater. In addition, at the crack tip, imperfections in the final

regions of fracture were observed. These types of fractures were caused by transition in the stress

state from plane-strain to plane-stress by means of reducing the cross-sectional area during crack

propagation. The effective crack length, which indicates a limit of data mining, was followed

0.5
1000λ Pmax
according to the ISO-11782-2 standard, such that aeff ≤ W − , where aeff is the
2 Bσ Y

effective crack length, W is the width of the specimen in the crack propagation direction,

λ = 4W , B is the depth of the specimen, Pmax is the maximum load and σ Y is the yield

strength of the material. Thus, an adequate effective crack length was established for every test.

Additively, Figure 2(c) shows some rust on the surface of the specimen due to exposure to seawater.

3.1 Fatigue crack propagations of the base metal and HAZs

5
 Figures 3(a), 3(b) and 3(c) show the experimentally obtained FCP and CFCP behaviors of

HSB800 in air and seawater in terms of the FCPR and CFCPR due to the stress intensity factor

(SIF) range under given loading conditions. As can be seen, the CFCPRs are dramatically increased

in seawater, as compared to the FCPRs in air, for any loading conditions. However, the levels of

da/dN between the FCPR and the CFCPR in the same loading conditions varied according to the

material. As can be seen, the acceleration of the CFCPR for the base metal was higher than the

HAZs. For example, at the same SIF range of K=60 MPa·m1/2, the CFCPR was 41 times larger

than the FCPR for the base metal, but the CFCPRs for CG and ICCG-HAZs were only 7.7 and 7.3

times larger than the respective FCPRs.

Specifically, as shown in Figure 3(a) for the base metal, the behaviors show that regions I

and II could be clearly indentified in both air and seawater environments. Region III, however, was

not easily identified in the seawater condition. Furthermore, the threshold values of the SIF range

were increased with respect to the decreases in R and f in both air and seawater. The threshold

values of the SIF range were decreased 53% and 59% for a load condition of R=f=0.5 as compared

to the values for R=f=0.1 in air and seawater, respectively. On the other hand, as can be seen in

Figures 3(b) and 3(c), it was difficult to identify not only the region demarcations but also the

obvious threshold values of the SIF range.

In addition, the FCPR for base metal in air was smoothly accelerated in the transition

domains of regions II and III. For the cases of the FCPR and the CFCPR in CG-HAZ and ICCG-

HAZ, however, the accelerations were rapid and angulated. Similar results were achieved in the

seawater except for cases of base metal. Possible reasons are discussed in the following section.

3.2 Microstructures

Micro-photographs of the fracture surfaces of base metal, CG-HAZ and ICCG-HAZ in the

6
air condition are presented in Figure 4. Transgranular (TG) crack propagation occurred during the

fatigue test in the base metal. In addition, at the end of the fatigue crack region indicated in Figure

4(a), laminar cracks perpendicular to the propagation were produced due to the laminar

microstructures of the as-rolled plate material. Figures 4(b) and 4(c) show different fracture aspects

in the CG-HAZ and ICCG-HAZ, as compared to those of the base metal. As can be seen, the crack

propagation mechanism was changed from transgranular cracking to intergranular (IG) cracking. At

the transition region of the fatigue crack propagation, as shown in Figure 3, the FCPR and CFCPR

was suddenly accelerated. A possible reason for this could be that the CG-HAZ and ICCG-HAZ

have relatively larger grain sizes than does the base metal, and they have isotropic microstructures

due to the heat-induced grain growth. An increased grain size would change the crack propagation

mechanism from TG to IG in a ferritic steel as found by Lopes and Charlier [12]. In addition,

Serebrinski et al. [13] figured out that a higher level of stress caused by higher strain rate could lead

out the transition of crack propagation mechanism from TG to IG. Meanwhile, it was recognized

that the FCPR by IG fracture was much more higher than the FCPR by TG fracture [14,15]. In this

work, thus, the transition of mechanisms (TG to IG) in HAZs, as shown in Figure 4(b) and 4(c),

may occur by the relatively larger grain size together with increasing stress intensity factor during

the progress of fatigue fracture. Simultaneously, the slope of increasing FCPR and CFCPR were

suddenly accelerated as can be seen in Figure 3(b) and 4(c). Figure 5 show the SEM photographs of

1,000x magnifications for the materials. Laminar crack in base metal and intergranular cracks in

CG- and ICCG-HAZ were shown in Figure 5(a), 5(b) and 5(c), respectively. Consequently, the

sudden accelerations in the FCPR and the CFCPR in both the CG- and ICCG-HAZs shown in

Figure 3 may have been caused by the transition in the crack propagation mechanism from TG

cracking to IG cracking, as shown in Figure 4 and 5. Optical microscopy and SEM photographs

were obtained by NeoMaterials Hybrid Process Laboratory, Hanyang University.

7
4. Conclusions

The FCP and the CFCP behaviors of HSB800 were investigated both in air and seawater

environments. The CFCPRs are dramatically increased in seawater as compared to the FCPRs in air

for every case of loading condition. However, the levels of da/dN between the FCPR and the

CFCPR in the same conditions varied according to the materials. The acceleration of the CFCPR in

seawater was dominant for the base metal, in which both regions I and II are obvious in both air

and seawater environments. Region III, however, was not easily identified in the seawater condition.

Furthermore, the threshold values of the SIF range were increased with respect to the decreases in

R and f in both air and seawater solution. It has been shown that a transition in the fracture

mechanisms from transgranular to intergranular cracking in HAZ materials can lead to sudden

accelerations in the FCPR and CFCPR during the fracture process.

Acknowledgements

This work was supported by the Research Institute of Industrial Science and Technology.

References

[1] M.L. Schmidt, F.D. McGuire, H.E. Lu, in: 36th Mechanical Working and Steel Processing,

Proceedings, Baltimore, Maryland, USA, 16-19 Oct. 1994, Publ. Iron and Steel Society/AIME,

Warrendale, PA, USA, 1994, pp. 215-229.

[2] R.I. Murakami, Y.H. Kim, W.G. Ferguson, Contribution of the cyclic loading portion below the

opeing load to fatigue crack growth, Fatigue Fract. Eng. Mater. Struct. 14 (1991) 741-748.

8
[3] D. Kujawski, A new ( K+ Kmax)0.5 driving force parameter for crack growth in aluminum alloys,

Int. J. Fatigue 23 (2001) 733-740.

[4] K. Sadananda, A.K. Vasudevan, Crack tip driving forces and crack growth representation under

fatigue, Int. J. Fatigue 26 (2004) 39-47.

[5] P.K. Liaw, A.W. Logsdon, Fatigue crack growth behavior in inconel 706 at 297K and 4.2K, Acta

Metall. 36 (1988) 1731-1744.

[6] J.E. King, Enhanced model of partial crack closure for correlation of R-ratio effects in

aluminum alloy, Fatigue Fract. Eng. Mater. Struct. 5 (1982) 177-188.

[7] H.S. Khatak, in: A.S. Kanna, M.K. Totlani, S.K. Singh (Eds.), Corrosion fatigue problems in

ferritic steels and austenitic steels, Corrosion and its control (CORCON'97), Proceedings, Dec. 3-6,

vol. 1, Elsevier, 1997, pp. 249-257.

[8] S. Sivaprasad, S. Tarafder, V.R. Ranganath, M. Tarafder, K.K. Ray, Corrosion fatigue crack

growth behaviour of naval steels, Corrosion Sci. 48 (2006) 1996-2013.

[9] M. Itatani, J. Fukakura, M. Asano, M. Kikuchi, N. Chujo, Fatigue crack growth behavior of

weld heat-affected zone of type 304 stainless steel in high temperature water, Nuclear Eng. Design

153 (1994) 27-34.

[10] M.E. Somervuori, M.T. Alenius, T. Kosonen, R. Karppi, H.E. Hanninen, Corrosion fatigue of

weld-bonded austenitic stainless steels in 3.5% NaCl solution, Mat. Corrosion 57/7 (2006) 562-567.

[11] M. Akita, M. Nakajima, K. Tokaji, T. Shimizu, Fatigue crack propagation of 444 stainless steel

welded joints in air and in 3%NaCl aqueous solution, Mat. Design 27 (2006) 92-99.

[12] L.C.R. Lopes, J. Charlier, Effect of grain size and integranular stresses on the cyclic behaviour

of a ferritic steel, Mat. Sci. Eng A 169 (1993) 67-77.

[13] S.A. Serebrinsky, G.S. Duffo, J.R. Galvele, Effect of strain rate on stress corrosion crack

velocity: difference between intergranular and transgranular cracking, Corr. Sci. 41 (1999) 191-195.

9
[14] D. Tromans, R.-H. Sun, Intergranular/transgranular fatigue of copper: influence of

environment on crack path and propagation rates, Mat. Sci. Eng. A 219 (1996) 56-65.

[15] L.H. Burck, J. Weertman, A comparison of yield strength effects on brittle transgranular ad

intergranular fatigue crack growth rates, Scripta Mat. 12/5 (1978) 417-420.

10
Figures and Tables

Figure 1. Schematic diagram showing the experimental apparatus for the CFCP test.

(a) (b)

(c)
Figure 2. Photographs showing the (a) initial and final specimens in (b) air and (c) seawater.

11
(a)

12
(b)

13
 (c)
Figure 3. Fatigue and corrosion fatigue crack propagation behaviors of HSB800 under several
loading conditions in air and seawater environments; (a) base metal, (b) CG-HAZ, and (c) ICCG-
HAZ.

14
 (a) (b)

(c)
Figure 4. Photographs showing the optical microscopic observations of 25x magnifications of the
fracture surfaces in air for (a) base metal, (b) CG-HAZ, and (c) ICCG-HAZ.

15
 (a) (b)

(c)
Figure 5. Photographs showing the SEM observation of 1,000x magnifications for
(a) base metal, (b) CG-HAZ, and (c) ICCG-HAZ.

16
Table 1. Chemical composition of HSB800
C Si Mn P S Cu Cr Ni
0.10 0.65 2.20 0.015 0.006 0.10-0.50 0.45-0.75 0.05-0.80

Table 2. Mechanical properties of HSB800

YS UTS E elongation
690 MPa 800 MPa 205 GPa 16%

Table 3. Summary of test conditions

load frequency, f
0.1 0.3 0.5
load ratio, R
Base metal air & seawater
0.1 CG-HAZ air & seawater
ICCG-HAZ air & seawater
Base metal air & seawater
0.3 CG-HAZ air & seawater
ICCG-HAZ air & seawater
Base metal air & seawater
0.5 CG-HAZ air & seawater
ICCG-HAZ air & seawater
*The actual test conditions are indicated by darkened boxes

17