Study of fatigue crack growth behaviour of Al-Mg and Al-Mg-Sc-Zr alloys

Abstract

Fatigue crack growth behaviour of Al-Mg and Al-Mg-Sc-Zr alloys is studied under different orientations and stress ratios. The microstructures and the fracture morphologies of each alloy are observed. For both alloys, they have Brass, S and Goss texture; while only Al-Mg alloy has Cube texture which indicates recrystallisation; as fatigue crack propagated, intergranular fracture transformed into transgranular fracture accompanied with gradually coarser crack surfaces. Al-Mg alloy is featured by large recrystallised grain, favourable plasticity, rougher crack surfaces, and with relatively strong crack growth resistance before rapid growth. While Al-Mg-Sc-Zr alloy with small grain size and relatively smoother crack surfaces, shown cyclic stability at higher ΔK level.

Keywords

Al-Mg Al-Mg-Sc-Zr fatigue crack growth fracture morphology

Introduction

In modern engineering, most structural components are susceptible to alternating stress-induced fatigue problems. Especially as mechanisation is accelerated and modern industries like aerospace develop continuously, various components or machinery operate in harsher mechanical environments, where accidents resulting from fatigue failures account for a considerable proportion. Hence, the study of the fatigue crack growth (FCG) behaviours of pre-cracked specimens can provide significant information of the fatigue fracture of materials, and provide a material selection guide for the components that may be subjected to external damage and lead to stress concentration.

An effective method of improving the alloy strength is grain refinement. With annealing after deformation, the grain size can be reduced. Among microstructural characteristics, grain boundaries are known as the primary obstacles for short crack propagation. The decelerated crack development at grain boundaries can be attributed to the decrease of strain amplitude at the crack tip due to the resistance to the slip bands. As the grain size increases, the FCG rate decreases with the increases of the threshold value, which is ascribed to the grain size-related roughness effect of fracture surface [1]. When the crack is smaller than the grain size, it exhibits slowly or no propagation [2]. Fatigue failure mode will be transformed into the Paris state when the size of the plastic deformation zone is equivalent to the characteristic microstructure size of the alloy [3]. High-tenacity and low-threshold alloys are recommended for low-cyclic fatigue under high stress, while high-threshold alloys are suggested for high-cyclic fatigue under low stress [4,5].

Crack closure exists in Al alloys, while the plastic behaviours at the crack tip will be changed significantly [6]. Grain size and fracture roughness can also affect the crack closure and deflection [7]. In general, higher ΔK_th and corresponding slower propagation can be achieved for coarser grain structures, associated with higher roughness and closure [8]. The residual stress will influence the FCG behaviours and change the material response to the dynamic loads [9]. Beneficial residual compressive stresses can impede crack initiation and growth [10]. The growth of large and small cracks is associated with a plasticity-induced crack closure model [11]. The near-ΔK_th growth of small cracks relies upon the microstructural characteristics of alloys. The crack closure effect is no longer apparent in FCG stages II and III.

Changes in crack propagation mechanisms are associated with the increase of the dimension of the plastic zone and the type/quantity/morphology of the damaged structural characteristics within [12,13]. With the interaction of the crack tip with the local shear zone, interdendritic boundary, particle/material interface, shrinkage void, and particle fracture [14], and the enhanced shielding effect of the crack tip due to crack branching and deflection, the FCG rate is decelerated [14]. The connection-induced crack branching was observed from the AA 2324-T39 alloys under different directions [15]. Such a deviation in the crack path presents when ΔK is high, which, however, is not affected by the stress ratio [16]. Owing to the crack branching diffusion, the increasing numbers of interactively propagating cracks and grain boundaries as well as the crack-precipitate interaction, ultrafine grained 7075 Al alloys had stronger anti-FCG ability than that of coarse-grained alloys, especially at high ΔK level [17]. The FCG rate of alloys will be decelerated with the fall in temperature due to the reduced hydrogen embrittlement [18] and humidity [19]. Irregular fracture surfaces may restrain the water vapour from moving towards the crack tip, thus resulting in a weaker temperature effect [19].

Al-Mg alloy offered higher tensile ductility, more even deformation behaviours, and stronger anti-macrocrack growth ability [20], while Al-Mg-Sc-Zr alloy with coherent Al₃(Sc, Zr) precipitates, and rather refined sub-grain structures, offered higher fatigue strength [20]. Although the addition of Sc, Zr significantly improves the yield strength, tensile strength [20,21] and microhardness [21] of Al-Mg alloy, the effect on anti-FCG ability is not necessarily improved. In this study, the FCG behaviours of unmodified and Sc, Zr modified Al-Mg alloys were explored through FCG tests. Both alloys were subjected to FCG tests at R = 0.1 and 0.8. The specimen has two orientations, for the T-L specimen, crack propagates along the rolling direction, and for the L-T specimen, crack propagates along the transverse direction. Meanwhile, the fatigue crack micro-characteristics of each specimen in different FCG stages were compared under SEM, to study the possible causes of the difference in crack growth performance between the two alloys. It provides a reference for material selection in aeronautical structural components, which may suffer from fatigue crack propagation problems.

Materials and methods

Northeast Light Alloys Ltd. provided the raw materials used in the experiments and the nominal composition information of the alloys. The cold-rolled Al-Mg (UA) and Al-Mg-Sc-Zr (MA) alloy sheets were annealed at 320°C for 1 h to stabilise the structure. The chemical composition of both alloys is listed in Table 1. The MA alloy contains 0.25 wt-% of Sc and 0.1 wt-% of Zr. Standard C(T) specimens were processed for the FCG tests performed using MTS 810 test platform at the stress ratios of R = 0.1 and R = 0.8 in a room temperature atmosphere environment following ASTM E647 [23]. Specimens with different orientations are shown in Figure 1. The test setup is illustrated in Figure 2(a). The sinusoidal load was used at a frequency of 10 Hz. After pre-crack, the average load and the load amplitude were kept constant until the specimen fractured, and the crack length was measured simultaneously using the compliance method and visual method, as shown in Figure 2(b).

Table 1

Main chemical composition of alloy sheets (wt-%).

C	Sc	Zr	Mn	Mg	Al
Al-Mg	—	—	0.4	5.8–6.0	Bal
Al-Mg-Sc-Zr	0.25	0.1	0.4	5.8–6.0	Bal

The calculation method of this test is based on ASTM E647 [23]. For each standard C(T) specimen, the stress-intensity factor range, ΔK, at the crack tip is a function of force range, ΔP, specimen thickness, B, sample width, W, and crack size a. It is expressed in Equation (1): (1) where α = a/W. The data that satisfy the following condition was considered to be valid and was selected for further analysis and comparisons. (2)

Metallographic specimens were prepared by electrolytic polishing with anodic coating. Before electrolytic polishing, the surfaces of the specimens were polished into a complete plane with 400# silicon carbide sandpaper, followed by polishing with 800# metallographic sandpaper until there were no obvious scratches. The chromic acid solution was used for electrolytic polishing. The polishing voltage was 22–24 V, and the current was controlled at about 0.6 mA. Fluoroboric acid was used in the coating solution. The coating voltage was 17 V at a coating current of less than 0.6 mA. The coating time was 1 ∼ 3 min. The fracture morphologies and precipitation of each alloy were observed using SEM and EDS. D8 Discover X-ray diffractometer was used to test the texture of electrolytic polished samples. The texture was measured by Cu target, tube voltage was 40 KV, tube current was 40 mA, and the measurement step was 5 degrees. Three incomplete pole figures were measured, and the texture results were represented by orientation distribution function (ODF) cross-section of ϕ2.

Results

Metallographic test results

The obtained three-dimensional metallographic results are shown in Figure 3. The grains of the two alloys were elongated along the rolling direction after the rolling process. Both surfaces show a typical layered structure, with each layer parallel to the rolling surface. In the UA alloy, a typical recrystallisation phenomenon was observed with larger grain size, more complete granularity, and thicker grain layers. In contrast, no recrystallisation was observed in the MA alloy with thinner grain layers. The grains were severely deformed along the rolling direction and were fibrous and pancake-like.

EDS test results

SEM analysis in Figure 4(a and b) show that there is a large number of micron-sized second-phase particles dispersed along the rolling direction in the UA alloy. The corresponding EDS results of each point are shown in Table 2. EDS analysis indicates that these particles were composed of Al, Fe, and Mn. The quantitative analysis indicates that they are the Al₆(Fe, Mn) particles. Figure 4(c and d) present some primary second-phase Al₃(Sc, Zr) particles and Al₆(Fe, Mn) particles in the MA alloy.

Table 2

EDS analysis results of UA (Spot A, B), and MA (Spot C, D) alloys.

	Spot A		Spot B		Spot C		Spot D
	wt-%	at.-%	wt-%	at.-%	wt-%	at.-%	wt-%	at.-%
Al	69.14	82.21	70.04	82.81	65.54	79.68	59.78	78.21
Sc	—	—	—	—	—	—	15.65	12.29
Zr	—	—	—	—	—	—	24.56	9.5
Fe	23.4	13.44	21.71	12.40	26.17	15.37	—	—
Mn	7.46	4.35	8.26	4.79	8.3	4.95	—	—

Textures

Pole and ODF figures of the textures of UA and MA alloys are shown in Figures 5 and 6, respectively. The volume fraction of each texture of the two alloys is shown in Table 3. It can be seen from Figure 5 that there are mainly Brass texture, Cube texture, Goss texture and S texture in UA alloy, among which Brass texture has the most content and S texture has the least content. Cube texture is usually considered as a typical recrystallisation texture. Therefore, it can be further confirmed that the recrystallisation behaviour does occur in UA alloy. Figure 6 shows that the main texture types in MA alloy are S texture, Brass texture, Goss texture and Copper texture, among which S texture content is the most, Copper texture content is the least, and basically no Cube texture appears.

Table 3

Volume fraction of each texture of UA and MA alloys.

Vol.-%	Brass	S	Goss	Cube	Copper
Al-Mg	26.7	5.84	9.29	10.62	—
Al-Mg-Sc-Zr	14.57	25.62	13.25	—	6.32

FCG test results

The FCG test results under two stress ratios of both alloys are displayed in Figures 7 and 8, respectively and the da/dN–ΔK values are shown in Tables 4 and 5. All the da/dN-ΔK plots demonstrated a typical three-stage crack growth process. For both alloys, the ΔK value was significantly reduced at R = 0.8. For the UA alloy, the anti-FCG ability of the L-T specimen was slightly stronger than that of the T-L specimen at both stress ratios. However, for the MA alloy, such anisotropy is not clear at R = 0.1. While it presents the abnormal anisotropy with the stronger anti-FCG ability from the T-L rather than the L-T specimen at R = 0.8. The UA alloy showed stronger anti-FCG ability in FCG stages I and II, but it presented lower ΔK values in FCG stage III. At R = 0.1, the ΔK value at the beginning of FCG stage III from the MA alloy was about 22 MPa·m^0.5, which is much higher than that of the UA alloy (about 16 MPa·m^0.5). A similar trend was also observed at R = 0.8.

Table 4

ΔK values of the two alloys at different da/dN levels (R = 0.1).

	ΔK (MPa·m^1/2)
	Al-Mg		Al-Mg-Sc-Zr
da/dN (mm/cycle)	L-T	T-L	L-T	T-L
1 × 10⁻⁵	5.294	4.875	4.767	4.573
1 × 10⁻⁴	8.955	8.446	6.934	7.071
4 × 10⁻⁴	16.524	15.361	14.689	12.646
1 × 10⁻³	22.491	20.314	23.159	21.226
4 × 10⁻³	/	/	31.679	31.826

Table 5

ΔK values of the two alloys at different da/dN levels (R = 0.8).

	ΔK (MPa·m^1/2)
	Al-Mg		Al-Mg-Sc-Zr
da/dN (mm/cycle)	L-T	T-L	L-T	T-L
2 × 10⁻⁵	4.302	4.135	3.566	3.747
4 × 10⁻⁵	4.923	4.794	4.033	4.507
7 × 10⁻⁵	5.463	5.304	4.683	5.197
1 × 10⁻⁴	6.543	6.378	5.407	6.336
3 × 10⁻⁴	/	/	9.037	9.459

The crack lengths of each specimen were measured through the compliance method and visual method during FCG. Table 6 shows the results from the UA alloy while Table 7 summarised the results from the MA alloy. Initially, the crack lengths of the UA alloy measured by the compliance method were longer than that by the visual method. Nevertheless, the difference was gradually narrowed with the increase in crack length. At R = 0.8, the values measured by the two methods differed very little. For the MA alloy, the values obtained by the two methods were rather similar.

Table 6

a₀ and a values of UA alloy in the FCG process.

R = 0.1, L-T		R = 0.1, T-L		R = 0.8, L-T		R = 0.8, T-L
a ₀	a	a ₀	a	a ₀	a	a ₀	a
30	30.462	31	31.416	30	30.187	30	30.083
40	40.442	40	40.420	35	35.184	35	35.014
50	50.133	50	49.877	40	40.146	40	39.977
66	66.012	65	64.983	45	45.052	44	43.919

Table 7

a₀ and a values of MA alloy in the FCG process.

R = 0.1, L-T		R = 0.1, T-L		R = 0.8, L-T		R = 0.8, T-L
a ₀	a	a ₀	a	a ₀	a	a ₀	a
30	30.022	31	31.199	34	34.205	30	30.226
40	40.067	49	49.023	40	40.209	40	40.273
54	54.083	62	62.007	50	50.122	50	50.188
70	70.057	75	74.946	60	60.102	60	59.962

Observation of crack growth path

The crack growth paths of the MA alloy were observed under the optical test platform, as shown in Figure 9. The initial da/dN of the FCG was around 10⁻⁵ mm/cycle for both stress ratios. As the crack size increased, the ΔK also increased, which in turn raised the da/dN, and the fatigue crack propagated rapidly until the specimen fractures.

The crack path was relatively smooth, representing linear growth at first, but started deflection with observable fluctuations. Finally, zigzag cracks were observed with severe fluctuation. Similar behaviours were observed at the two different stress ratios. The crack path is generally longer and smoother when R = 0.1 (Figure 9(a)), while shorter and the zigzag fluctuations are more pronounced when R = 0.8 (Figure 9(b)).

SEM observation of fracture surfaces

The FCG fracture surfaces of the UA and the MA alloy specimens with T-L orientation in different FCG stages were observed via the SEM. Both alloys were subjected to a typical fatigue fracture, with their macro-fracture morphologies divided into crack growth zone and transient fracture zone.

Figure 10 shows the fracture surface in FCG stage I for specimens of both alloys with T-L orientation. For both alloys, fatigue cracks grew inward along with the main slip system in a shear form. The small plastic deformation zone at the crack tip was extended by several grain lengths. Under a small driving force in stage I, the cracks grew along the direction and path consuming the least energy, which is usually controlled by the grain. Hence, the fatigue crack was more inclined to grow along the grain boundaries on a relatively smooth path. From the aspect of fracture morphologies, both alloys showed typical intergranular fractures. As shown in Figure 10(a), the SEM image of the fracture surface from the UA alloy indicates an uneven fracture surface. Meanwhile, due to the small grain size, the stratified structures were denser in the MA alloy, shown in Figure 10(b), and the overall fracture was relatively flat. Moreover, a stratified structure formed by rolling was observed with a bamboo-like fibrous grain contour of the longitudinal section of the rolled sheet.

In the FCG stage II, the fracture morphologies of both alloys became rougher, especially for MA alloy comparing to stage I as shown in Figure 11(b). Despite the continued intergranular fracture characteristics of both alloys, the fracture surface was much rougher in this stage relative to the FCG stage I. This was because the plastic deformation zone was enlarged in stage II. This can exert a combined action with multiple grains with a wider and larger fracture surface than that in the FCG stage I, which can cause the fluctuation and deflection of the crack path. The intergranular fractures were reduced while the transgranular fractures started appearing. Because of lower yield strength, the UA alloy had a larger plastic deformation zone and suffered from more serious deformation, which had adverse effects on FCG.

The SEM images of the fractures in the FCG stage III are shown in Figure 12. The intergranular fracture morphology could not be observed. The impacts of grain size and grain-boundary morphology on the FCG were gradually mitigated, accompanied by the ever-increasing proportion of transgranular fractures. As the ΔK value was further increased in this stage, the driving force for FCG was also increased, resulting in the expansion of the plastic zone. Moreover, the grains at the crack tip generated serious plastic deformation. The FCG path was coarsened and lengthened accompanied by the evident fluctuation and deflection, and the energy consumption was increased. All these changes could obstruct the crack growth.

The typical fatigue fracture morphologies during FCG are exhibited in Figure 13, including the fatigue striations of the UA alloy in Figure 13(a) and the dimples of the MA alloy in Figure 13(b). The former emerged from the FCG stage II, which became more evident with the increase in the FCG rate, and disappeared in the final fracture zone. Fatigue striations in the UA alloy were continuously distributed in different fracture surfaces perpendicularly to the FCG direction. The dimple morphologies observed in the final fracture zone in the MA alloy represent the feature of ductile fracture which is similar to the fracture surface of static tension. The fracture was dominated by the transgranular fracture. The dimples were elongated towards the rolling direction, including small dimples and crushed grains.

Discussion

The FCG pattern showed differences in different FCG stages. In FCG stage I, da/dN was slow due to the lacking of driving force for crack growth, and the crack grew in a discontinuous manner. In stage II, the fatigue crack developed continuously with the increased driving force for FCG, while the fatigue striation occurred, a straight line (Paris zone) was observed in the coordinates for the da/dN-ΔK curve in stage II, as fatigue cracks were transited into the Paris stage, fracture behaviours were no longer sensitive to the microstructure. Ductile striations produced at intermediate-to-high ΔK occur by plastic blunting/re-sharpening process during loading/unloading. [32]. In stage III, the crack developed rapidly under a great FCG driving force, making it unstable through the cyclic loading, fatigue striations gradually disappeared and the dimples were formed in the final fracture zone. The grain structure can also influence the FCG behaviours of the Al alloys. The impeded dislocations are likely to be accumulated at the grain boundaries. As a result, the intergranular fracture occurred with the stress concentrated at the grain boundaries. Differences in grain direction affect the direction of fatigue crack propagation, making the crack path tortuous and hindering crack propagation. In addition, FCG behaviours can be impacted by structural characteristics like second phase characteristic, intermetallic particles, and defects generated by impurities as well as non-metallic inclusions.

Study shows that the larger the recrystallised grain size, the more energy is consumed and the slower the FCG rate [25]. In this study, it was concluded that the cracks grew along the grain boundaries in FCG stage I. Stratification was observed on the fractured surface with the spacing as same as the grain size. In FCG stage II, the crack started to gradually transform into transgranular growth with reduced influences from the grain structure. UA alloy featured larger recrystallised grains, rougher crack surfaces, and more tortuous crack paths. Results in strong anti-FCG ability when their cracks grew along the grain boundaries. MA alloy with smoother crack growth path and lower energy consumption for developing fatigue cracks because of its smaller sizes of grains and sub-grains. As a result, it has a relatively lower resistance to fatigue cracks. However, with the increased ΔK and the enlarged plastic zone, more grains were involved in the deformation and the sizes of individual grains had fewer influences on the FCG behaviour, the high-density grain boundaries and the coherent Al₃(Sc, Zr) precipitates allow MA alloy to exhibit better stability under high ΔK level, providing increased resistance to crack growth. Consequently, in FCG stage III at a high level of ΔK, the MA alloy had stronger anti-FCG ability with respect to the UA alloy.

The stress ratio also significantly affects the FCG behaviour. Higher stress ratio condition will result in higher average stress of the specimens with a higher crack growth driving force. A higher stress ratio lowers the FCG threshold, reduces the crack closure, and accelerates the FCG rate. For Al-Mg alloy with coarse grain structure, the increase in the stress ratio, the fatigue threshold was found to be reduced while the Paris slope was not affected [22]. The crack closure theory suggests that fatigue cracks can only grow under a completely opened state. Certain opening stress is required for the complete opening of cracks due to elastic deformation. Higher stress ratio may eliminate the crack closure effect especially in FCG stage I when ΔK value is low, because it leads to higher minimum stress-intensity factor K _min. According to ASTM E561[24], for C(T) specimens, the plastic zone size r_Y at the crack tip is a function of K and yield strength σ_YS , expressed as r_Y = K/(2π·σ_YS ), so for materials with lower yield strength, larger plastic zone would be formed at the same K level, plasticity induced crack closure effect on a higher level would hinder the FCG due to its greater consumption of crack growth energy.

UA alloy showed possible crack closure at R = 0.1 in FGC stage I. The crack surface was repeatedly contacted during the load cycle, the crack closure induced by plasticity and fracture roughness helped to prevent the FCG in the UA alloy. As the crack continued to grow with increased ΔK value, the crack closure became less effective. Since it was more likely to reach the opening stress at R = 0.8, no significant crack closure was observed in the UA alloy. The MA alloy had a smaller plastic zone with more concentrated stress due to its higher yield strength, showed no significant crack closure under both stress ratios. The UA alloy showed anisotropy at both R = 0.1 and R = 0.8, L-T specimens showed better anti-FCG ability. While MA alloy showed irregular anisotropy at R = 0.1 and anomalous anisotropy at R = 0.8 when T-L specimens showed better anti-FCG ability.

The value of the Schmid factor directly determines the difficulty of activating a slip system. The Schmid factor is defined as follows [28]: m = cosϕ·cosλ, where ϕ and λ are the intersection angles between the external load and the slip plane and the slip direction, respectively. Study illustrated that grains with higher Schmid factor presents easier crack growth [29] by minimising the energy expenditure for FCG [31]. If the crack tip is facing ‘hard’ orientation zone with lower Schmid factor, the difficulty of activating the slip system makes the main crack bypassing this area and grow within the ‘soft’ orientation zone with higher Schmid factor [26]. A higher stress ratio R can not only increase the FCG driving force, but also increase the proportion of grains with high Schmid factor, which is finally manifested as the increase of FCG rate [27].

Study reported that there are more Brass and S textures grains with larger Schmid factors in the RD compared to the TD, as the FCG rate in the RD is relatively higher than that in the TD [30]. Refining the grains with Goss orientation could induce more crack deflection, consequently enhancing the fatigue resistance [33]. Sheet with high-intensity Goss texture presents a lower FCG rate than relatively low intensity Goss-texture one. Greater crack deflections in the high intensity Goss texture sheet are observed during FCG, resulting in a rougher fracture surface and a greater roughness-induced crack closure effect [34]. Goss grain possesses the largest fatigue crack deflection, followed by Cube orientation, while Brass, Copper, S and Random grains have relatively small crack deflection angles [35].

In this study, Brass texture, S texture and Goss texture are found in both alloys. Brass texture and S texture are conducive to improving the anti-FCG ability in TD direction, and the existence of Goss texture is conducive to causing crack deflection and improving the anti-FCG ability. For UA alloy, Cube texture indicates recrystallisation, which is conducive to crack deflection and increase FCG resistance. The content of S texture in MA alloy is high, indicating that the content of internal subgrain structure is high. No obvious Cube texture was found, and recrystallisation was not obvious. There is also a certain amount of Copper texture in the alloy, which is beneficial to improve the FCG resistance in RD direction, which may be one of the reasons for the complexity of the anisotropy in MA alloy.

Conclusions

Fatigue crack growth tests of Al-Mg and Al-Mg-Sc-Zr alloys were carried out under different orientations and stress ratios. The metallographic, SEM, EDS, XRD analysis of both alloys and SEM observation of the fracture surfaces were carried out. The following conclusions were obtained:

Both alloys exhibit the change from intergranular fracture to transgranular fracture as the ΔK increases. For the Al-Mg alloy, crack closure is observed at R = 0.1, but it gradually disappears with the increase of ΔK. No obvious crack closure is identified in Al-Mg-Sc-Zr alloy under all stress ratios.

Al-Mg alloy has a stronger anti-FCG ability in FCG stages I-II due to its large grain size, high plasticity, larger plastic deformation zones, rough crack surfaces, and more tortuous growth paths, although it is instability under high stress intensity conditions.

Al-Mg-Sc-Zr alloy has weaker anti-FCG ability than the Al-Mg alloy because of its smaller grain size, higher grain boundary density, relatively smooth crack surface, and flat crack path for crack growth. However, Al-Mg-Sc-Zr alloy was resistant to instability under high stress intensity owing to the impacts of grain refinement and dispersion of strengthening.

For both alloys, there are Brass texture and S texture which can increase crack resistance in TD direction and Goss texture which can cause crack deflection. Significant recrystallisation was found in the Al-Mg alloy and L-T specimens had slightly stronger anti-FCG ability than T-L specimens. No significant recrystallisation in Al-Mg-Sc-Zr alloy and it has anomalous anisotropy under the high stress ratio.

Footnotes

Acknowledgements

The authors would like to express their gratitude to EditSprings () for the expert linguistic services provided.

Disclosure statement

No potential conflict of interest was reported by the author(s).

References

Pao

P S

Holtz

R L

Jones

H N

Effect of environment on fatigue crack growth in ultrafine grain Al–Mg. Int J Fatigue. 2009;31(11-12):1678–1683.

Lorenzino

Navarro

Growth of very long “short cracks” initiated at holes. Int J Fatigue. 2015;71:64–74.

Han

S W

Kumai

Sato

Fatigue crack growth behaviour in semi-liquid die-cast Al-7% Si-0.4% Mg alloys with fine effective grain structure. Mater Sci Eng A. 2001;308(1-2):225–232.

Lados

D A

Apelian

Fatigue crack growth characteristics in cast Al–Si–Mg alloys. Mater Sci Eng A. 2004;385(1-2):200–211.

Lados

D A

Apelian

Fatigue crack growth characteristics in cast Al–Si–Mg alloys. Mater Sci Eng A. 2004;385(1-2):187–199.

Yang

Zhang

Liu

Existence and insufficiency of the crack closure for fatigue crack growth analysis. Int J Fatigue. 2014;62:144–153.

Pao

P S

Jones

H N

Cheng

S F

Fatigue crack propagation in ultrafine grained Al–Mg alloy. Int J Fatigue. 2005;27(10-12):1164–1169.

Lados

D A

Apelian

Donald

J K.

Fatigue crack growth mechanisms at the microstructure scale in Al–Si–Mg cast alloys: mechanisms in the near-threshold regime. Acta Mater. 2006;54(6):1475–1486.

Lammi

C J

Lados

D A.

Effects of residual stresses on fatigue crack growth behaviour of structural materials: analytical corrections. Int J Fatigue. 2011;33(7):858–867.

10.

Lammi

C J

Lados

D A.

Numerical predictions and experimental measurements of residual stresses in fatigue crack growth specimens. Eng Fract Mech. 2011;78(6):1114–1124.

11.

Newman

Jr J C

Anagnostou

E L

Rusk

Fatigue and crack-growth analyses on 7075-T651 aluminum alloy coupons under constant- and variable-amplitude loading. Int J Fatigue. 2014;62:133–143.

12.

Lados

D A

Apelian

Jones

P E

Microstructural mechanisms controlling fatigue crack growth in Al–Si–Mg cast alloys. Mater Sci Eng A. 2007;468-470:237–245.

13.

Lados

D A

Apelian

Relationships between microstructure and fatigue crack propagation paths in Al–Si–Mg cast alloys. Eng Fract Mech. 2008;75(3-4):821–832.

14.

Chan

K S

Jones

Wang

Fatigue crack growth and fracture paths in sand cast B319 and A356 aluminum alloys. Mater Sci Eng A. 2003;341(1-2):18–34.

15.

Bao

Zhang

A link-up resulted fatigue crack branching in Al–Cu–Mg alloy. Int J Fatigue. 2016;92:459–469.

16.

Zhang

Bao

Fei

Load effects on macroscopic scale fatigue crack growth path in 2324-T39 aluminium alloy thin plates. Int J Fatigue. 2014;58:193–201.

17.

Das

Jayaganthan

Chowdhury

Fatigue behaviour and crack growth rate of cryorolled Al 7075 alloy. Mater Sci Eng A. 2011;528(24):7124–7132.

18.

Burns

J T

Gupta

V K

Agnew

S R

Effect of low temperature on fatigue crack formation and microstructure-scale propagation in legacy and modern Al–Zn–Mg–Cu alloys. Int J Fatigue. 2013;55:268–275.

19.

Goma

F A T

Larouche

Bois-Brochu

Effect of extrusion aspect ratio and test temperatures on fatigue crack growth behaviour of a 2099-T83 Al–Li alloy. Int J Fatigue. 2014;59:244–253.

20.

Roder

Wirtz

Gysler

Fatigue properties of Al-Mg alloys with and without scandium. Mater Sci Eng A. 1997;234:181–184.

21.

Lathabai

Lloyd

P G.

The effect of scandium on the microstructure, mechanical properties and weldability of a cast Al–Mg alloy. Acta Mater. 2002;50(17):4275–4292.

22.

Brosi

J K

Lewandowski

J J.

Delamination of a sensitized commercial Al–Mg alloy during fatigue crack growth. Scr Mater. 2010;63(8):799–802.

23.

ASTM E647 . Standard Test Method for Measurement of Fatigue Crack Growth Rates. [S].

24.

ASTM E561 . Standard Test Method for KR Curve Determination. [S].

25.

Zhang

The role of recrystallization on fatigue crack propagation of a high alloying Al-Zn-Mg-Cu alloy. Mater Lett. 2022;311:131564.

26.

Jian

Luo

Tang

Influence of microstructure on fatigue crack propagation behaviors of an aluminum alloy: role of sheet thickness. Eng Fract Mech. 2017;180:105–114.

27.

Wang

Liu

Investigation of fatigue crack propagation behavior of 5083 aluminum alloy under various stress ratios: role of grain boundary and Schmid factor. Mater Sci Eng A. 2020;773:138871.

28.

Schmid

Boas

(1950). Plasticity of crystals. e. schmid and w. boas. f. a. hughes & co. ltd. london. 1950. 353 pp. 222 illustrations. index. appendix. bibliography. 35s. net. Aeronautical Journal, 54(479), 353–719.

29.

Sun

Wang

Wei

Study on small fatigue crack initiation and growth for friction stir welded joints. Mater Sci Eng A. 2019;739:71–85.

30.

Shi

Pan

Low anisotropy of fatigue crack growth in Al-5.8 Mg-0.25 Sc. Int J Fatigue. 2019;125:170–178.

31.

Spangenberger

A G

Lados

D A

Coleman

Microstructural mechanisms and advanced characterization of long and small fatigue crack growth in cast A356-T61 aluminum alloys. Int J Fatigue. 2017;97:202–213.

32.

Lynch

Some fractographic contributions to understanding fatigue crack growth. Int J Fatigue. 2017;104:12–26.

33.

Xia

Liu

Bai

Coupling effect of grain structures and residual secondary phases on fatigue crack propagation behavior in an Al-Cu-Mg alloy. J Mater Eng Perform. 2021;30(4):2669–2679.

34.

Liu

Goss texture intensity effect on fatigue crack propagation resistance in an Al-Cu-Mg alloy. J Alloys Compd. 2018;730:318–326.

35.

Zhao

Wahab

M A

Ling

Fatigue crack propagation within Al-Cu-Mg single crystals based on crystal plasticity and XFEM combined with cohesive zone model. Mater Des. 2021;210:110015.