Effects of pitting corrosion on fatigue life of aluminum alloy Y12CZ based on initial discontinuity state
YU Da-zhao(郁大照) 1, CHEN Yue-liang(陈跃良)2, HU Jia-lin(胡家林)1, YANG Mao-sheng(杨茂胜)1
1. Naval Aeronautical Engineering Academy, Yantai 264001, China;
2. Qingdao Branch, Naval Aeronautical Engineering Academy, Qingdao 266041, China
Received 28 July 2006; accepted 15 September 2006
Abstract: Based on initial discontinuity state (IDS) of material, a preliminary analytical model was presented to evaluate the effect of interaction of pitting corrosion and fatigue loading on the residual fatigue life of aluminum alloy LY12CZ. A life prediction was carried out using constant and variable amplitude loading for various pitting corrosion levels, and the prediction agreed reasonably with the available test data. The results suggest that the combination of a pit and IDS can be treated as the initial crack size. Pitting corrosion causes a significant decrease in fatigue lives with small corrosion depths. But the effect of pit on fatigue life is gradually reduced with increasing pit size. A pit with a constant depth can be applied to the model for long exposure structure. A preliminary recommendation for the pit depth is about 1 mm for LY12CZ. At last the effect of multiple-site corrosion damage (MSCD) on fatigue life was also studied, and the result shows that MSCD can decrease substantially fatigue life compared with that of a single crack.
Key words: fatigue life; pitting corrosion; aluminum alloy; initial discontinuity state; multiple site corrosion
1 Introduction
Aluminum alloys are widely used in commercial and military aircrafts because of their favorable properties of low mass and high strength. The heterogeneities within the microstructure of aluminum alloys make them susceptible to corrosion. A small site of corrosion attack can propagate into a big defect and lead to an early catastrophic failure. From a practical point of view, it is important to develop a model to evaluate the effect of interaction of pitting corrosion and fatigue loading on the residual fatigue life of structure.
In the past, most research on pitting corrosion was focused on the test methods of pitting susceptibility, materials resistance to pitting, pitting prevention and removal technology, as well as qualitative pitting NDI technology. Most of the research was from standpoint of chemistry or materials science[1-6]. Models for corrosion pit development and for the initiation of fatigue cracks under corrosion fatigue conditions have been proposed and have some validity for many materials and corrosion conditions under which they were derived[7-11]. In Ref.[12] corrosion pitting damage has been quantified and related to the decrease of fatigue life of 2024-T3 specimens corroded in alternate immersion corrosion profile. In Ref.[13], corrosion pitting is considered responsible for the initiation and early propagation of short fatigue cracks. In Ref.[14], the effect of corrosion on multiple site damage scenarios and aircraft structural integrity is considered such as accounting for the onset of multiple site damage (MSD) from corrosion pits. On the other hand, MEDVED et al[15] concluded that not all of the pits will turn into fatigue cracks. To date, due to its complexity, there is still no general model available to predict the service life of corroded aluminum structures and design against fatigue in the presence of corrosion. A number of scholars have suggested that the initial discontinuity state (IDS) approach might be a route in which corrosion damage could be quantitatively assessed in fatigue design[15-16].
The objective of this study is to evaluate an analytical model which is developed to study the effect of interaction of pitting corrosion and fatigue loading on the residual fatigue life of aluminum alloy LY12CZ based on IDS. In addition, the results of this study highlight several other interesting aspects of pitting corrosion in fatigue.
2 Determination of IDS for LY12CZ aluminum alloy
The key element of the process proposed in this paper is the method of computing the IDS of the LY12CZ aluminum alloy. The initial condition provides some of the necessary input to fracture mechanic algorithms that could account for the statistical variation of life from time zero, thus capturing the inherent scatter observed in fatigue.
A crack growth analysis software AFGROW was used to extrapolate IDS from experimental results [17]. For a certain fatigue specimen with a known failure cycle numbers, an optimal IDS value which fits the test data best could be obtained. The method varied with IDS values to obtain a series of crack growth prediction, and then the prediction yielding the best correlation between the analysis and experimental data defined the IDS.
There were 91 specimens for which an IDS could be determined. The distribution of IDS thus was obtained. The mean IDS was 0.045 mm with a stand deviation of 0.223 on the flaw size.
3 Description of pitting corrosion model based on IDS
The analytical model of corrosion pits on the formation and growth of fatigue cracks is best understood by beginning with the model for fatigue crack formation and growth in the noncorroded state. In the as-fabricated, noncorroded, or pristine state, fatigue cracks in aluminum alloys have been found to form at constituent particles[18]. These particles on discontinuities in the alloy matrix are on grain boundaries, pores, etc. An initial discontinuity state (IDS) is defined based upon the discontinuities present in the pristine state. The IDS concept is similar to Equivalent Initial Flaw Size (EIFS) concept, but it is coupled to those physical material features which play decisive roles in structural failures. So compared with EIFS, the term IDS goes a long way in theory, and its distribution represents the inherent condition of the as-manufactured material.
Corrosion pits at the same constituent particles form. Thus, for the case of corrosion followed by fatigue in a benign environment, the analysis starts using the pit dimensions plus IDS to account for adjacent constituent particles. As the pit grows, it does not consume the crack but the local stress amplification changes.
Based on discussion above, the refined model used in this paper is shown in Fig.1. For simultaneous corrosion and fatigue, the analysis begins with the crack growth model and the pit growth model acting on the surface discontinuity. The local stress amplification due to the pit is included in the crack growth model.
Fig.1 Schematic diagram of pitting corrosion developing process
4 AFGROW model
4.1 Damage locations
The mapping of damage in structure shows that pitting typically occurs around fastener holes. For modeling, efforts should be focused on the regions around fastener holes. Fig.2 shows the three simplified locations where cracking could be seen to develop if exfoliation, intergranular corrosion, or pitting attack occurs around a countersunk fastener hole.
Fig.2 General locations for pitting around fastener hole
Case 2 represents a severe stress condition compared with the other two locations and has been incorporated into AFGROW[19], so the following sensitivity study uses the geometry—a single corner crack at the bore of a straight-shank fastener hole, as shown in Fig.3.
Fig.3 AFGROW models for two scenarios crack growth analysis
At the same time, the corrosive environment may also accelerate the onset of multiple-site corrosion damage (MSCD) in aluminum alloy that contains the same structural elements such as fastener holes [20-21]. So the study also models the double corner cracks at a hole to evaluate the effect of MSCD on fatigue life, as shown in Fig.3 (b).
4.2 Model inputs
The model was assumed to be 127 mm wide and 4.6 mm thick based on a typical wing skin thickness. The NASGRO material model for LY12CZ aluminum was used along with constant and variable amplitude spectrum. The maximum stress of constant amplitude is 220 MPa, R=0.1. For variable amplitude loading the Fighter Aircraft Loading Standard For Fatigue (FALSTAFE) spectrum was selected.
IDS was treated as a semi-circular surface crack within AFGROW in current analysis. To approximate pitting, cracks of aspect ratio equal to one were added to the IDS size.
5 AFGROW analysis results
5.1 Validity of process zone model based on IDS
In order to actually simulate the pitting corrosion process, the sizes of pit at different exposure time were the same as the measured test data [15]. Fig.4 displays experimental results and prediction of fatigue life based on AFGROW model. As can be seen from Fig.4, the model gives a good prediction of the exfoliation life. It shows that the model used in current analysis is suite to analyze the effect of pitting corrosion on fatigue life for both constant and variable amplitude spectrum.
5.2 Corrosion damage size effects
The depth of pitting attack found at the hole was evaluated at three depths: 0.050, 0.125, and 0.254 mm. Fig.5 shows the results of the AFGROW simulation and depicts interesting trends. The plots show that the significant decrease in fatigue life as a result of pit development followed by a very gradual reduction in fatigue life with increasing pit size is observed. So an approximate, but reasonable, pitting-affected life estimation can be obtained by assuming a pit with a constant depth and then applying the model described in this paper. Based on the data in current analysis, a pit depth about 1mm would be an appropriate value. This indicates that knowing the rate at which pits grow is not like knowing when corrosion pits first form. However, additional work at other stress levels is needed to confirm this viewpoint.
Fig.4 Comparison of experimental lives with predicted results
Fig.5 Impact of initial discontinuity size (IDS + pit) and MSCD on fatigue life
5.3 MSCD cracks effects
To evaluate the effect of MSCD cracks, the double same corner cracks at a hole were modeled for two spectrum types.
Fig.5 also compares the results for the model (a) and (b) shown in Fig.3. The plot shows a significant decrease in fatigue life for the case of diametrically opposed cracks compared with that of a single crack, as observed in test. But this decreased trend is not sensitive to spectrum type.
6 Conclusions
A model to assess the effects of pitting corrosion on fatigue life was proposed. A limited set of fatigue life data from tests demonstrates that this model is reasonable. Fatigue lives are obtained for different corrosion damage sizes and the case in presence of MSCD cracks.
1) For the case of corrosion followed by fatigue in a benign environment, the analysis start can be treated as the pit dimensions plus IDS to account for adjacent constituent particles.
2) The development of a corrosion pit causes a significant drop in fatigue life, but the structural fatigue life does not continue to decrease proportionately with continuous corrosion.
3) A pit with a constant depth can be applied to the model described in this paper. A preliminary recommendation for the pit depth is about 1mm.
4) The effect of MSCD is to decrease substantially fatigue life compared with that of a single crack. But this trend is not sensitive to spectrum type.
References
[1] ROKHLI S L, KIM J Y, NAGY H. Effect of pitting corrosion on fatigue crack initiation and fatigue life[J]. Eng Fract Mech, 1996, 62: 425.
[2] CHEN G S, GAO M, WEI R P. Microconstituent induced pitting corrosion in a 2024-T3 aluminum alloy [J]. Corrosion, 1996, 52(1): 8-15.
[3] GAO M, FENG C R, WEI R P. An AEM study of constituent particles in commercial 7075-T6and 2024-T3 alloys [J]. Metallurgical and Material Transactions A, 1998, 29: 1153-1160.
[4] LIAO C M, OLIVE J M, WEI R P. In suite monitoring of pitting corrosion in 2024 aluminum alloy [J]. Corrosion, 1998, 54(6): 451-458.
[5] ABOLIKHINA E V, MOLYAR A G. Corrosion of aircraft structures made of aluminum alloys [J]. Materials Science, 2003, 39: 889-894.
[6] WU X R, LI B, LU F, TAO C H, ZHANG W F, XI N S. Analysis and control of corrosion cracking in airframe structures [J]. Engineering Failure Analysis, 2006, 13: 398-408.
[7] HARLOW D G, WEI R P. A probability model for the growth of corrosion pits in aluminum alloys induced by constituent particles [J]. Eng Fract Mech, 1997, 59: 305-325.
[8] HARLOW D G, WEI R P. Probability modeling and statistical analysis of damage in the lower wing skins of two retired B-707 aircraft [J]. Fatigue Fract Engng Mater Struct, 2001, 24: 523-535.
[9] DUQUESNAY D L, UNDERHILL P R, BRITT H J. Fatigue crack growth from corrosion damage in 7075-T6511 aluminum alloy under aircraft loading [J]. In J Fatigue, 2003, 25: 371-377.
[10] ZHOU S, TURNBULL A. Influence of pitting on the fatigue life of a turbine blade steel [J]. Fatigue Fract Engng Mater Struct, 1999, 22: 1083-1093.
[11] DOLLEY E J, LEE B, WEI R P. The effect of potting corrosion on fatigue life [J]. Fatigue Eng Mater Struct, 2000, 23: 555-560.
[12] ZAMBER J E. HILLBERRY B M. Probabilistic approach to predicting fatigue lives of corroded 2024-T3 [J]. AIAA J, 1999, 37: 1311-1317.
[13] PYUM S, SEUNG J O, SOO W N. Corrosion fatigue crack initiation of Al-Zn-Mg-Mn alloy in 0.5 M Na2SO4 solution [J]. Mater Sci Eng A, 1997, 241: 281-284.
[14] WEI R P, HARLOW D G. Corrosion-Enhanced fatigue and multiple-site damage [J]. AIAA J, 2003, 41(10): 2045-2050.
[15] MEDVED J J, BRETON M, IRVING P E. Corrosion pit size distributions and fatigue lives—A study of the EIFS technique for fatigue design in the presence of corrosion [J]. In J Fatigue, 2004, 26: 71-80.
[16] FAWAZ S A. Equivalent initial flaw size testing and analysis of transport aircraft skin splices [J]. Fatigue Fract Engng Mater Struct, 2003, 26: 279-290.
[17] HU J L. Research on Fatigue Life Prediction Methods of Aircraft Aluminum Alloy Based on IDS [D]. Shandong: Naval Aeronautical Engineering Institute, 2006.
[18] NEWMAN R P, PHILLIPS E P. Fatigue analysis under constant- and variable-amplitude loading using small crack theory[R]. NASA TM-209329, 1999.
[19] HARTER J A. AFGROW Users Guide and Technical Manual, AFGROW for Windows 2K/XP [M].Version 4.0009. 2004.
[20] BELLINGER N C, KOMOROWSKI T J, BENAK T J. Residual life predictions of corroded fuselage lap joints [J]. In J Fatigue, 2001, 23: s349-s536.
[21] BELLINGER N C, KOMOROWSKI T J, GROULD R W. Corrosion pillowing in aircraft fuselage lap joints[J]. AIAA 2005-2023.
(Edited by YANG Bing)
Foundation item: Project(50675221) supported by the National Natural Science Foundation of China
Corresponding author: YU Da-zhao; Tel: +86-532-88033160; E-mail: ytyudazhao@yahoo.com.cn
