- Email: [email protected]
Effects of crack orientation on the fatigue crack growth rate and fracture toughness of AA6063 alloy deformed by ECAP
Materials Science & Engineering A 733 (2018) 71–79
Contents lists available at ScienceDirect
Materials Science & Engineering A journal homepage: www.elsevier.com/locate/msea
Effects of crack orientation on the fatigue crack growth rate and fracture toughness of AA6063 alloy deformed by ECAP Mohammad Ali Kazemi, Rahman Seifi
T
⁎
Department of Mechanical Engineering, Faculty of Engineering, Bu-Ali Sina University, P.O.C 65175-4161 Hamedan, Iran
A R T I C LE I N FO
A B S T R A C T
Keywords: ECAP Fracture toughness Fatigue crack growth Crack orientation
In this study, effects of crack plane orientation on the fatigue crack growth rate and fracture toughness of 6063 aluminum alloy deformed by equal channel angular pressing (ECAP) were investigated. The ECAP process continued up to five passes without failure. Grain refinement was obvious after five passes of the ECAP process. Textural studies showed aligning the grains in known directions. After four passes, yield and ultimate strengths increase respectively from 90 MPa and 209 MPa to 300MPa and 375 MPa and also reduction in elongation was observed. The roughness decreased after the process. The fatigue crack growth rate was investigated at different load ranges with the same load ratio for different orientations. The crack growth rate increased after one pass of the ECAP process. After five passes, the AA6063 shows a lower crack growth rate in compared with as-received material. The fracture toughness of mode I and mixed-mode for different orientations were measured. The results showed that the orientation has a significant effect on the fatigue crack growth and fracture toughness of the ECAPed samples. The fracture surfaces were studied using scanning electron microscope (SEM) and refined equiaxed dimples were observed after the ECAP process.
1. Introduction
movement plays an important role in the mechanical properties changes after this process. The fatigue behavior of severely deformed metals was investigated by Vinogradov and Hashimoto [10]. The results showed that the grain boundaries have important effects on both low and high cycle fatigue behaviors of these materials. The SPD materials have shorter low cycle fatigue(LCF) lives in compared with coarse-grained solids because of large decrease in ductility during the process. However, the high cycle fatigue (HCF) lives increased after the SPD process. The shear banding and crack nucleation strongly depends on the material and type of the SPD process. An optimum balance between strength and ductility leads to better fatigue properties [11]. The effects of the ECAP process with subsequent thermal mechanical treatments on the HCF strength of titanium were investigated by Semenova et al. [12]. They found that Low-temperature annealing of samples increased the endurance limit, strength and ductility. Namdar and Jahromi [13] studied the effects of the ECAP on the fatigue behavior of AA2011 alloy. The experimental results revealed that the HCF enhanced by increasing the strength of the material [14]. While the enhancement of tensile strength due to the grain refinement is sensible, the fatigue strength of the UFG materials is not improved in the same manner. The effect of the grain refinement, texture and chemical composition of the material on the fatigue behavior of
The average grain size plays a crucial role in the mechanical properties of a material. Different methods of severe plastic deformations (SPD) were used to produce ultrafine grained (UFG) materials with improved properties such as strength, hardness, fatigue life and fracture toughness. The ECAP method is the common and simple process between the SPD processes. The ECAP process has wide applications for different metallic alloys. This process increased the strength of the material by imposing high strains to introduce the high density of dislocations [1]. The cross-section area of the specimen has no significant changes after the ECAP process. Different routes are available for samples to pass through intersecting channels such as A, B and C routes [2]. The route A is the most widely used route. In this route as can be seen in Fig. 1, the specimen has no rotation according to previous pass. The UFG materials has wide applications in energy (oil-gas), biomedical, aerospace, sports, etc. [3]. Many researchers have studied mechanical properties and microstructures of this materials such as aluminum [4–6] and Mg alloys [7,8]. Horita et al. [9] studied the effects of the ECAP process on the mechanical properties of the aluminum alloys. They found that the strength increased after the ECAP process, however elongation to failure showed a large decrease after the first pass. The dislocation ⁎
Corresponding author. E-mail address: rseifi@basu.ac.ir (R. Seifi).
https://doi.org/10.1016/j.msea.2018.07.042 Received 30 May 2018; Received in revised form 7 July 2018; Accepted 12 July 2018 0921-5093/ © 2018 Elsevier B.V. All rights reserved.
Materials Science & Engineering A 733 (2018) 71–79
M.A. Kazemi, R. Seifi
Fig. 1. Fundamental routes used in ECAP process.
metallic alloys makes it difficult to investigate the influence of only one of these parameters on the fatigue performance. The chemical composition plays crucial role in the fatigue behavior of light alloys [15]. Cavaliere [16] studied the fatigue characteristics of pure metal (Al, Ti, Ni and Cu) produced by the ECAP process. The experimental results showed that the fatigue crack growth rate increased in the pure UFG metals in compared with CG (coarse grained) ones. The fatigue tests revealed that the ECAPed material is less sensitive to the crack initiation due to decreasing the average grain size. Vinogradov et al. [17] investigated the effect of the ECAP process on the fatigue crack growth in copper samples. They found that for small values of ΔKI (stress intensity factor range for mode I), the crack propagation rates for ECAPed copper is higher than that of the CG copper. The growth rate is the same for the UFG and CG materials at higher ΔKI regime. Fatigue crack initiation life for fine grain brass H62 was investigated by Zheng et al. [18]. They found that the finer grains create higher strength in the brass and cause intergranular cracking under the fatigue loadings. Effects of the ECAP process on the fracture toughness of the materials are studied by Seifi and Kazemi [5] for AA6063, Yu et al. [19] for AZ31 and Rahmatabadi et al. [20] for pure AA1050. There are a few studies conducted in field of the effect of the crack plane orientation on the fatigue and fracture behavior of the UFG materials. Hohenwarter and Pippan [21] studied the fracture toughness of the ECAP-deformed iron and effects of the crack plane orientation on the fracture toughness. They examined the fracture toughness in different orientations and measured the critical fracture toughness, KIC and critical J integral, JIC . The results showed that the crack plane orientation has a significant effect on the fracture toughness. Hohenwarter and Pippan [22] investigated the fracture toughness of nanopolycrystalline metals produced by the SPD process. They studied the recent results about the fracture properties of different UFG (d < 1 µm) and nanocrystalline (d < 100 nm) metals. In this study, the effects of the crack plane orientation on the fatigue crack growth of AA6063 aluminum alloy fabricated by the ECAP process have been investigated experimentally. The results were compared with the fatigue and fracture properties of the CG as-received alloy. The microstructure, texture and strength of the ECAPed aluminum were evaluated. The SEM images were taken from fracture surfaces of the samples (ECAPed and CG) to determine the behaviors.
Fig. 2. The cross section of the assembled die.
die was used for the ECAP process as depicted schematically in Fig. 2. The die material was H13 tool steel. This type of the die is suitable for reducing the time of the process, because it doesn’t have bolts. The channels intersecting angle is Φ = 90° and outer corner angle is Ψ = 22°. In order to achieve better grain refinements and mechanical properties, rout A was selected for conducting the ECAP process based on recommendation in [24]. The ECAP process applied to billets up to five passes at 200 °C. The pressing speed of 0.5 mm/s was applied to specimens during the process. Equivalent strain is approximately 1.05 in every pass based on Eq. (1) which presents equivalent strain, εN after N passes of the ECAP process [1]. The strains that imposed on the sample are large simple shear [25].
εN =
N ⎡ Φ+Ψ Φ+Ψ ⎤ ⎞ ⎞ + Ψ cosec ⎛ 2 cot ⎛ 3⎣ ⎝ 2 ⎠⎦ ⎝ 2 ⎠
(1)
The specimens for tensile tests, Compact Tension (CT) and mixedmode samples in different orientations were machined from ECAPed billets. A billet after one pass is shown in Fig. 3. The cross-sectional shape of the billet after extrusion was square. The tensile tests were carried out by using a zwick/roell universal testing machine according to ASTM E8 standard [26]. The tensile specimens were machined from the center of the billets as depicted in Fig. 3(c) with dimensions as follows: gauge length= 45 mm, radius of fillets= 8 mm and diameter= 9 mm. Three tests were performed for each condition. The strain rate and crosshead speed applied on the samples are 3.7 × 10−4s−1 and 1 mm/min, respectively. The microstructure, texture and fracture surface studied by using scanning electron microscope (SEM), atomic force microscope (AFM) and x-ray diffraction (XRD). The standard CT specimens were made from as-received and the ECAPed materials in different orientations according to ASTM E647 [27] for fatigue crack growth and fracture toughness measurements based on ASTM E1820 standard [28]. As depicted in Fig. 4, different types of the CT samples were made and noted as L, M and N. In L type, the crack surface is normal to normal direction (ND) and will grew in extrusion direction (ED), while in M and N types, the crack is normal to the ED. The growth direction of the cracks in M and N types are in the ND and transverse direction (TD), respectively. The width and thickness of the CT specimens were considered as W = 10mm and B = 3mm , respectively. The electric discharge machining (EDM) method was used to create sharp notches in the samples. In order to conduct the fatigue cyclic loading, the servo-hydraulic test
2. Experimental work We used 6063 aluminum alloy (AA6063) for preparing the sample tests. The chemical composition of AA6063 is given in Table 1. The asreceived specimens were annealed at 420 °C for 3 h before ECAP process to achieve better deformability and reduce friction due to the decrease of dislocation density [23]. MoS2 lubricant was used to decrease the friction. Specimens with dimensions of 12 × 12 × 110 mm (Fig. 3a) in form of billets were machined from aluminum rods with 20 mm diameter. A two pieces split Table 1 Chemical composition (wt%) of AA6063. Al
Si
Fe
Cu
Mn
Mg
Cr
Ni
Zn
Ti
Pb
Co
Bi
Base
0.44
0.93
3.50
0.59
1.02
0.03
0.005
0.38
0.04
0.10
0.005
0.004
72
Materials Science & Engineering A 733 (2018) 71–79
M.A. Kazemi, R. Seifi
Fig. 5. a) Schematic view of the modified Arcan fixture and assembled specimens in the fixture b) α = 15° c) α = 60°.
fI = 5.266 − 32.736β + 99.968β 2 − 127.078β 3 + 65.331β 4 fII = 0.3657 + 1.646β − 0.083β 2 Fig. 3. (a) schematic view of As-received Billet, (b) billet after one pass of ECAP, (c) schematic view of tensile sample machined from the center of billet.
3. Results and discussions 3.1. Microstructure and textural studies Fig. 6 shows the microstructure of as-received and ECAPed materials after different passes by using the SEM and AFM techniques. The grain size for as-received aluminum alloy was calculated in the range of 8–75 µm with averaged size as 45 µm. As can be seen, after first pass, the grain size decreased. The crystallite size for the ECAPed materials were calculated using the XRD profile and Williamson-Hall method. This method is valid for the grain size less than 2 µm [32]. The obtained results were confirmed in comparing with work done by Majzoobi et al. [33]. It was assumed that both the crystallite size and the grain size have reduction after the process. The crystallite size after five passes of the ECAP process was measured as 230 nm. The intensity values (in a.u.) for different 2θ degrees are depicted in Fig. 7. The roughness measurements were performed by AFM on the lateral cross-section of as-received and ECAPed specimens. Fig. 8 showed 3D image of the specimens’ surfaces for as-received and five times ECAPed material. The image showed more homogeneous roughness for ECAPed material. The average roughness values were 140 nm (SD=8.4) and 123 nm (SD=7.12) for as-received and ECAPed materials, respectively. Also, pole figures of [2 2 0] and [1 1 1] planes of as-received and ECAPed samples are shown in Fig. 9. The texture of the ECAPed samples depends on the as-received material texture until first pass, and for further passes texture more depends on ECAP process routes. The cubic lattice structures allow slippage to occur more easily than non-cubic lattices, so hcp metals are not as ductile as the fcc metals [34]. There are some deviations textures that departure from the ED, which may be caused by the refinement of the grains. The basal planes are not rotated because of using rout A in the ECAP process. [35]. The results showed that Cθ is the strongest component of texture for specimen after first passes of the ECAP, however, for further passes, Bθ / B′θ is the strongest component. The selected route for ECAP process plays an important role in the texture of the material for further passes of the process [34]. The basal planes in [2 2 0] were oriented parallel to the extrusion direction after five passes. Maximum texture intensity after five passes increased from 5.345 to 10.624 for [2 2 0] plane and from 5.61 to 7.06 for [1 1 1] plane.
Fig. 4. a) CT specimen b) schematic views of L, M and N types.
machine (zwick/roell) was used. The tests were performed at frequency as 30 Hz and load ratio was R = 0.1 for different load ranges. The high resolution camera was used for measuring the crack length at different life times. According to ASTM E647 standard for CT specimens, the stress intensity factor range was calculated by using the following equation:
ΔK =
ΔP f (β ) f (β ) = (2 + β )[0.886 + 4.64β − 13.32β 2 + 14.72β 3 (BBN W )1/2 − 5.6β 4]/(1 − β )3/2
(2)
Where β = a/ W and also ΔP , a, W , B and BN are the load range, crack length, sample width, thickness and net thickness for the samples with side grooves, respectively. If side grooves are not present, then BN = B . For plain strain condition, according to ASTM standard, the sample thickness must be greater than 2.5 mm [29]. The CT and CTS (CT shear) specimens were used for calculation of the fracture toughness under pure modes and mixed-mode loading conditions, respectively. The fatigue pre-crack was created at the end of 4 mm notch. The fracture load-displacement curves were recorded as output of the tests. The ASTM standards E1820 and D5045 [30] were considered. Determination of the mixed-mode fracture toughness has some difficulties due to limitation in the billets dimensions, so Arcan fixture was modified to apply the proper loads on the samples in different angles, as depicted in Fig. 5. The values of mixed-mode stress intensity factors were calculated by employing the following equations [31]:
KI =
P cos α πa Psinα πa f (β ), KII = f (β ), Kt = 1.25BW I 1.25BW II
KI 2 + KII 2
3.2. Tensile strength
(3)
The stress-strain curves for as-received and ECAPed samples are shown in Fig. 10. Young's modulus is about E = 70 GPa and other
KI , KII and Kt are mode I (opening), mode II (sliding mode) and mixed-mode stress intensity factors, respectively. Where: 73
Materials Science & Engineering A 733 (2018) 71–79
M.A. Kazemi, R. Seifi
Fig. 6. SEM images of a) as-received sample (x500), b) as-received (x1800), c) after first pass with x2000, AFM images of TD-ND surfaces d) as-received e) after five passes.
tensile properties given in Table 2. The ECAPed UFG material shows enhancement of dislocations density so the strength would be increased [1]. The remarkable enhancement of tensile and yield strengths occurred after first pass. After the first pass, these values improved by 150% and 45%, respectively while after that until fourth pass (during three passes) improvements are about 233% and 79%, respectively with respect to as-received samples. After the first pass, there are considerable reduction in ductility of the alloy (about 39%), as the results of Horita et al. [9] and after fourth passes, elongation decreased about 58%. 3.3. Fatigue crack growth Under constant range (or amplitude) of the applied stress, based on the relation ΔK = YΔσ πa , stress intensity range ( ΔK ) is proportional with the crack length. For small values of crack length and so small ΔK , the crack propagation is sensitive to grain size. For UFG material, the crack has to break more boundaries because the grain's boundaries are closer to each other [36]. For greater values of ΔK , crack propagation obey a power law manner [37] noted as Paris equation and is less sensitive to grain size.
Fig. 7. Intensity values versus 2θ for a sample after five passes.
74
Materials Science & Engineering A 733 (2018) 71–79
M.A. Kazemi, R. Seifi
Fig. 8. AFM 3D images of the surfaces, a) as-received, b) after five passes.
Fig. 9. Pole figures for as-received and ECAPed AA6063 alloy.
Fig. 11. Fatigue crack length for as-received and ECAPed material at R= 0.1 and Pmax= 200 N.
growth values of as-received and ECAPed materials for M type specimens with R= 0.1 and Pmax= 200 N are shown in Fig. 11. The crack growth rates were increased after first pass and so the life time decreased. The main reason is high decreasing of the elongation and embrittling after first pass [9]. After five passes, AA6063 alloy shows a lower crack growth rate in compared with as-received material. As can be seen, growth rates are low for small cracks in all three cases. Rate increasing is continued up to the different crack lengths. One can see that for as-received sample, crack growth rate increases up to about a= 1.5 mm, while for sample with one pass to about a= 1.0 mm and for five passes to a= 0.75 mm. These behaviors show that with increasing the ECAP passes, the growth rate reaches to its final value in small crack length with more retarding cycles. After these crack lengths, slope of aN curves (growth rates) are almost constant with the same values. Fig. 12 shows the fracture surfaces of the CT specimens for one pass ECAPed material with different sample orientations at R= 0.1 and Pmax = 150 N. There are some minor differences between images. The fracture surfaces of M and N types (Fig. 12(b) and (c)) had the same crack plane orientations. The fracture surface for M and N types specimens indicated more similarity in compared with L type. The fatigue crack growth for one pass ECAPed alloy at different crack plane orientations with R= 0.1 and Pmax= 150 N loading is depicted in Fig. 13. The results show that the crack plane orientation has a significant effect on the crack growth rate. The M and N types specimens have lower rate in compared with L type. The crack growth rate for M type is close to N type because of the same crack plane orientation. The textural study and fracture surfaces in Fig. 12 confirmed the obtained results. Fig. 14 shows the fatigue crack length versus life time for as-received and ECAPed alloy with R= 0.1 at different load ranges. The results reveal that the fatigue life enhanced significantly by decreasing
Fig. 10. Stress-strain curves of AA6063 before and after the ECAP process. Table 2 Tensile properties for as-received and ECAPed samples (all data are the mean values of three tests). sample
Yield Strength (MPa)
Ultimate Strength (MPa)
Elongation (%)
As-received After one pass After four passes
90 225 300
209 304 375
16.6 10.0 7.0
Eq. (4) presents Paris law:
da = C (ΔK )m dN
(4)
where a and N are the crack length and number of the fatigue cycles, respectively. C and m are material parameters [38]. The fatigue crack 75
Materials Science & Engineering A 733 (2018) 71–79
M.A. Kazemi, R. Seifi
Fig. 12. Fracture surfaces for one pass ECAPed samples under Pmax= 150 N and R= 0.1. a) L type, b) M type, c) N type.
Fig. 15. Fatigue crack behavior of Al-6063 after one pass ECAP for different load range at R = 0.1.
maximum load from Pmax= 200 N to Pmax= 150 N. Fig. 11 shows that the maximum effect of the grain refinement under Pmax= 200 N is at beginning of the fatigue crack growth step and for crack length more than 1.5 mm, the three curves of the fatigue crack growth were approximately parallel. As can be seen in Fig. 14 under the load Pmax= 150 N, effect of the grain refinement continued up to a= 1.75 mm. The obtained results reveal that for small load ranges (Pmax=150 N), the effects on the growth rate is increased up to a= 1.75 mm for five passes of the ECAP and up to a= 1.35 mm after one pass. For higher load ranges (Pmax=200 N) effect of the refinement decreased up to a= 0.5 mm for five passes. These results confirmed by other researches [10]. The da/dN versus ΔK at R= 0.1 with different load ranges for the ECAPed alloy is shown in Fig. 15. The fatigue crack behavior of asreceived and five passes ECAPed material with the same load range is depicted in Fig. 16. The fatigue crack growth in Paris regime has significant changes for aluminum alloy before and after ECAP process and is less sensitive to load range. Similar results for effect of the load range on the fatigue crack growth in Paris regime obtained by other researchers [38,39]. The Paris equation parameters, C and m were calculated for
Fig. 13. Fatigue crack length for one pass ECAPed material with different sample orientation at R= 0.1 and max= 150 N.
Fig. 14. Fatigue crack growth for different sample orientations at R= 0.1 and different load ranges. Fig. 16. Fatigue crack behavior of Al-6063 before and after ECAP process at R= 0.1. 76
Materials Science & Engineering A 733 (2018) 71–79
M.A. Kazemi, R. Seifi
Table 3 Paris law parameters for different types of ECAPed AA6063 alloy. ECAPed sample (pass 1)
m
C
L type
5.34
10−12
M type
4.03
10−10
N type
3.98
10−10
different specimen orientations. The results are given in Table 3. As can be seen, the parameters are different for various orientations. The m values increase from 4.03 and 3.98 for M and N types to 5.34 for L type and also C decreases from 10−10 to 10−12 with metric dimensions for rate and SIF values in Paris equation. The M and N types have the same fatigue crack growth behavior while L type sample has higher value.
Fig. 17. Normalized first mode fracture toughness in different passes of different types.
3.4. Fracture toughness
the sample. The loads were applied to the specimens in directions with angles as α = 15° and 60°, where α is the angle between load influence line and notch surface normal vector, as depicted in Fig. 5. The average values of the critical fracture loads were used to determine critical mixed-mode stress intensity factor. The values of unwanted mode III stress intensity factor is negligible due to the perfect symmetry of modified Arcan fixture [40]. A sample load curve and compliance variations lines for mixedmode loading at α = 60° is presented in Fig. 18. These curves were used in determination of the crack driving force which used in Eq. (3) for determining the SIFs in modes I and II. Mixed-mode fracture toughness, Kt was calculated with the above procedure. One of the loading angles (α = 15° ) is close to the opening mode and the other loading angle (α = 60°) is near to the shear loading mode. The fracture behavior of the material in mixed-mode varied independently from mode I for different types, because strength is a function of all types as L, M and N as listed in Table 5. Fig. 19 shows that the normalized fracture toughness, Kt / Kt0 for ECAPed samples and as-received material under various loadings. For α = 15° loading, the results are close to the values of fracture toughness for the pure mode (I) and the same behavior was expected.
The fracture toughness KIC was calculated with mentioned procedure in the ASTM standard (Eq. (2)) and then converted to critical J‐integral, JIC , by employing the following equation [28]:
JIC ≃
KIC 2 (1 − ν 2) E
(5)
The results for KIC and JIC values are presented in Table 4 for asreceived and extruded samples in different passes. After first pass of the ECAP process, the KIC values decreased from 18.4 MPa m for as-received material to 13.1, 15.71 and 14.43 MPa m for L, M and N types, respectively. The major reason of this reduction is decreasing the ductility of the material after the process in comparison with strength improvement. After further passes of the ECAP process, the ductility were decreased with lower rate and strength increased in each passes. Variations of normalized SIFs for mode I versus as-received material (KI / KI0 ) in different types are shown in Fig. 17. KI0 is the fracture toughness of the as-received material. The M and N specimen types showed less reduction of fracture toughness in compare with L type, which was similar to results obtained by Hohenwarter and Pippan [21]. In M and N types, the crack surfaces are normal to ED (Fig. 4) and according to the textural variations, little ductility increasing in these directions is expected. However, in L type, the surfaces are parallel to extrusion and SIFs have more reduction. The results show that crack plane orientation has crucial role in the fracture toughness of the material. For mixed-mode fracture toughness similar to mode I, we considered the crack growth condition as 5% increasing in compliance of
3.5. Fracture surface In order to study the effects of the ECAP process on the fracture surfaces of the samples, SEM images were taken. As can be seen in Fig. 20(a), large and deep dimples due to ductile fracture observed in as-received material. Fig. 20(b) shows the fracture surface of the ECAPed material. As can be seen, the number of dimples increases due to grain refinement after the ECAP process. Fine and shallower dimples are observed in compared with as-received material due to decreasing of the ductility which can be attributed to slow movement of
Table 4 J-integral and fracture toughness values for different passes of the ECAP process and different orientations (each value is mean value of three tests). Pass number As-received 1st
2nd
3rd
4th
5th
type
JIC (kJ / m2)
KIC (MPa m )
L M N L M N L M N L M N L M N
4.38 2.22 3.18 2.69 2.40 3.31 3.00 2.36 3.25 2.83 2.70 3.85 3.59 2.96 4.66 4.17
18.4 13.1 15.7 14.4 13.5 15.9 15.1 13.4 15.7 14.7 14.3 17.1 16.5 15.0 18.8 17.8
Fig. 18. A sample load curve and determination of mixed-mode crack driving force. 77
Materials Science & Engineering A 733 (2018) 71–79
M.A. Kazemi, R. Seifi
Table 5 Fracture toughness and J integral in mixed-mode loading for as-received and ECAPed materials (each value is mean of three tests). Pass number
(α ) degree
Type
KI MPa m
KII MPa m
Kt MPa m
J kJ m−2
As-received
15 60 15
L M N L M N L M N L M N
14.13 7.16 6.28 8.96 7.88 5.83 5.50 6.13 8.86 15.45 9.31 5.89 5.94 6.60
1.71 5.82 0.78 1.12 1.06 4.73 4.47 5.60 1.13 1.94 1.17 4.78 4.82 5.36
14.23 9.22 6.33 9.03 7.95 7.51 7.09 8.31 8.93 15.57 9.39 7.59 7.65 8.51
2.61 1.09 0.51 1.05 0.81 0.73 0.64 0.89 1.03 3.13 1.13 0.74 0.75 0.93
1st pass
60
5th pass
15
60
Fig. 19. mixed-mode fracture toughness for different types under various loadings.
Fig. 20. SEM images of the fracture surfaces: a, d) as-received b) after first passes of the ECAP process c) fracture zones.
4. Conclusion
dislocations in the UFG materials. In Fig. 20(c), two fracture zones: the fatigue crack growth and ductile fracture were specified. Fig. 20(d) shows the fracture surface of as-received material. The fatigue crack deflected along grain boundaries so the fracture surface shows fissured fracture.
Effects of the ECAP process on the fatigue crack growth and fracture behaviors of AA6063 alloy were investigated. Variations of the yield and ultimate strengths, fatigue crack growth rate, pure and mixed mode fracture toughness, texture, roughness, grain refining and fracture 78
Materials Science & Engineering A 733 (2018) 71–79
M.A. Kazemi, R. Seifi
surfaces were analyzed. The following conclusions can be obtained from this study:
[13]
1. Grain refinement was obvious after five passes of the ECAP process. The average grain size reduced from 45μm to about 230nm . 2. The yield and ultimate strength increased significantly after four passes from 90 MPa and 209 MPa to 300MPa and 375 MPa, respectively. The major increasing was after first pass and elongation decreased after four passes from 16.6% to 7%. 3. The roughness decreased for UFG material in compared with CG one. 4. The crack growth rate increases after one pass which is attributed to large decrease of the elongation. After five passes, AA6063 alloy shows a lower crack growth rate in compared with as-received material. The strength after five passes increased while the elongation decreased with small rate up to four passes. 5. The obtained results showed that the crack plane orientation had a significant effect on the crack growth rate and fracture toughness. The M and N types, because of the same crack plane orientation, had lower crack growth rate in compared with L type specimen. The Paris parameters were different for various orientations. 6. The obtained results revealed that for small load ranges, effects of the grain refinement on improvement of fatigue behavior increased and for higher load ranges decreased. 7. The fracture toughness for pure mode I decreased from 18.4 MPa m to 15.71 MPa m for M type samples. The fracture toughness increased after the third pass due to increase in the strength of the material in comparison with as-received one. 8. After the ECAP process fine and shallower dimples were observed in compared with as-received material.
[14]
[15]
[16] [17]
[18] [19]
[20]
[21] [22]
[23]
[24]
[25] [26]
Data availability statement
[27]
The raw/processed data required to reproduce these findings cannot be shared at this time as the data also forms part of an ongoing study.
[28] [29]
References
[30]
[1] R.Z.L. Valiev, G. Terence, Principles of equal-channel angular pressing as a processing tool for grain refinement, Prog. Mater. Sci. 51 (7) (2006) 881–981. [2] M. Gzyl, A. Rosochowski, S. Boczkal, L. Olejnik, The role of microstructure and texture in controlling mechanical properties of AZ31B magnesium alloy processed by I-ECAP, Mater. Sci. Eng.: A 638 (2015) 20–29. [3] Y. Huang, T.G. Langdon, Advances in ultrafine-grained materials, Mater. Today 16 (3) (2013) 85–93. [4] H. Darban, B. Mohammadi, F. Djavanroodi, Effect of equal channel angular pressing on fracture toughness of Al-7075, Eng. Fail. Anal. 65 (2016) 1–10. [5] R. Seifi, M.A. Kazemi, An investigation on the effects of equal channel angular pressing on the mixed‐mode fracture toughness and mechanical properties of 6063 aluminium alloy, Fatigue Fract. Eng. Mater. Struct. (2018), https://doi.org/10. 1111/ffe.12816. [6] E.K. Cardoso, V. Guido, G. Silva, W. Botta Filho, A. Jorge Junior, Microstructural evolution of AA7050 al alloy processed by ECAP, Matéria (Rio de Janeiro) 15 (2010) 291–298. [7] L. Tang, Y. Zhao, R.K. Islamgaliev, C.Y.A. Tsao, R.Z. Valiev, E.J. Lavernia, Y.T. Zhu, Enhanced strength and ductility of AZ80 Mg alloys by spray forming and ECAP, Mater. Sci. Eng.: A 670 (2016) 280–291. [8] L.B. Tong, M.Y. Zheng, X.S. Hu, K. Wu, S.W. Xu, S. Kamado, Y. Kojima, Influence of ECAP routes on microstructure and mechanical properties of Mg–Zn–Ca alloy, Mater. Sci. Eng.: A 527 (16–17) (2010) 4250–4256. [9] Z. Horita, T. Fujinami, M. Nemoto, T.G. Langdon, Improvement of mechanical properties for Al alloys using equal-channel angular pressing, J. Mater. Process. Technol. 117 (3) (2001) 288–292. [10] A. Vinogradov, S. Hashimoto, Fatigue of severely deformed metals, Adv. Eng. Mater. 5 (5) (2003) 351–358. [11] Z.-Z. Hu, Optimum balance between strength and ductility for fatigue properties, Eng. Fract. Mech. 48 (3) (1994) 445–453. [12] I.P. Semenova, R.Z. Valiev, E.B. Yakushina, G.H. Salimgareeva, T.C. Lowe, Strength
[31]
[32]
[33]
[34]
[35]
[36] [37] [38]
[39] [40]
79
and fatigue properties enhancement in ultrafine-grained Ti produced by severe plastic deformation, J. Mater. Sci. 43 (23) (2008) 7354. M. Namdar, S.A.J. Jahromi, Influence of ECAP on the fatigue behavior of agehardenable 2xxx aluminum alloy, Int. J. Miner., Metall., Mater. 22 (3) (2015) 285–291. Y. Hong, C. Sun, The nature and the mechanism of crack initiation and early growth for very-high-cycle fatigue of metallic materials – an overview, Theor. Appl. Fract. Mech. 92 (Supplement C) (2017) 331–350. Y. Estrin, A. Vinogradov, Fatigue behaviour of light alloys with ultrafine grain structure produced by severe plastic deformation: an overview, Int. J. Fatigue 32 (6) (2010) 898–907. P. Cavaliere, Fatigue properties and crack behavior of ultra-fine and nanocrystalline pure metals, Int. J. Fatigue 31 (10) (2009) 1476–1489. A. Vinogradov, T. Kawaguchi, Y. Kaneko, S. Hashimoto, Fatigue crack growth and related microstructure evolution in ultrafine grain copper processed by ECAP, Mater. Trans. 53 (1) (2012) 101–108. M. Zheng, M.X. Tong, H.P. Cai, C.Z. Xu, M.Q. Huang, Fatigue crack initiation life of fine grain brass H62, Theor. Appl. Fract. Mech. 54 (2) (2010) 105–109. X. Yu, Y. Li, L. Li, Fracture mechanism of AZ31 magnesium alloy processed by equal channel angular pressing comparing three point bending test and tensile test, Eng. Fail. Anal. 58 (2015) 322–335. D. Rahmatabadi, R. Hashemi, B. Mohammadi, T. Shojaee, Experimental evaluation of the plane stress fracture toughness for ultra-fine grained aluminum specimens prepared by accumulative roll bonding process, Mater. Sci. Eng.: A 708 (2017) 301–310. A. Hohenwarter, R. Pippan, Fracture of ECAP-deformed iron and the role of extrinsic toughening mechanisms, Acta Mater. 61 (8) (2013) 2973–2983. A. Hohenwarter, R. Pippan, Fracture and fracture toughness of nanopolycrystalline metals produced by severe plastic deformation, Philos. Trans. R. Soc. A: Math., Phys. Eng. Sci. 373 (2038) (2015). G.D. Fan, M.Y. Zheng, X.S. Hu, C. Xu, K. Wu, I.S. Golovin, Effect of heat treatment on internal friction in ECAP processed commercial pure Mg, J. Alloy. Compd. 549 (Supplement C) (2013) 38–45. K.J. Kim, D.Y. Yang, J.W. Yoon, Microstructural evolution and its effect on mechanical properties of commercially pure aluminum deformed by ECAE (Equal Channel Angular Extrusion) via routes A and C, Mater. Sci. Eng.: A 527 (29–30) (2010) 7927–7930. A. Vinogradov, S. Yasuoka, S. Hashimoto, On the effect of deformation mode on fatigue: simple shear vs. pure shear, Mater. Sci. Forum 584–586 (2008) 797–802. A. International, ASTM E8/E8M - 09 Standard Test Methods for Tension Testing of Metallic Materials, ASTM, 2009. A. International, ASTM E647-15e1, Standard Test Method for Measurement of Fatigue Crack Growth Rates, West Conshohocken, 2015. A. International, ASTM E1820-13 Standard Test Method for Measurement of Fracture Toughness, West Conshohocken, 2013. Y. Li, B. Gong, M. Corrado, C. Deng, D. Wang, Experimental investigation of out-ofplane constraint effect on fracture toughness of the SE(T) specimens, Int. J. Mech. Sci. 128–129 (Supplement C) (2017) 644–651. A. International, ASTM D5045-14, Standard Test Methods for Plane-Strain Fracture Toughness and Strain Energy Release Rate of Plastic Materials, West Conshohocken, 2014. N. Hallbäck, N. Jönsson, T-stress evaluations of mixed mode I/II fracture specimens and T-effects on mixed mode failure of aluminium, Int. J. Fract. 76 (2) (1996) 141–168. Y. Zhong, D. Ping, X. Song, F. Yin, Determination of grain size by XRD profile analysis and TEM counting in nano-structured Cu, J. Alloy. Compd. 476 (1–2) (2009) 113–117. G.-H. Majzoobi, J. Nemati, M.K. Pipelzadeh, S. Sulaiman, Characterization of mechanical properties of Al-6063 deformed by ECAE, Int. J. Adv. Manuf. Technol. 84 (1) (2016) 663–672. Mh Shaeri, M. Shaeri, Mt Salehi, S.H. Seyyedein, Mr Abutalebi, Texture evolution of ultrafine grained Al-7075 alloy produced by ECAP, Metall. Eng. 17 (56) (2015) 49–57. D. Hou, T. Liu, D. Shi, H. Chen, H. Chen, Study of twinning behaviors of rolled AZ31 magnesium alloy by interrupted in situ compressive tests, Mater. Sci. Eng.: A 653 (2016) 108–114. M.D. Sangid, H.J. Maier, H. Sehitoglu, The role of grain boundaries on fatigue crack initiation – An energy approach, Int. J. Plast. 27 (5) (2011) 801–821. X. Cai, R. Xia, M. Huo, J. Xu, A threshold formula for fatigue crack growth with mean stress intensity factors, Int. J. Mech. Sci. 135 (2018) 639–645. J. Horky, G. Khatibi, D. Setman, B. Weiss, M.J. Zehetbauer, Effect of microstructural stability on fatigue crack growth behaviour of nanostructured Cu, Mech. Mater. 67 (2013) 38–45. T. Zhao, J. Zhang, Y. Jiang, A study of fatigue crack growth of 7075-T651 aluminum alloy, Int. J. Fatigue 30 (7) (2008) 1169–1180. A.Eh Oskui, N. Choupani, M. Shameli, 3D characterization of mixed-mode fracture toughness of materials using a new loading device, Lat. Am. J. Solids Struct. 13 (2016) 1464–1482.
Contents lists available at ScienceDirect
Materials Science & Engineering A journal homepage: www.elsevier.com/locate/msea
Effects of crack orientation on the fatigue crack growth rate and fracture toughness of AA6063 alloy deformed by ECAP Mohammad Ali Kazemi, Rahman Seifi
T
⁎
Department of Mechanical Engineering, Faculty of Engineering, Bu-Ali Sina University, P.O.C 65175-4161 Hamedan, Iran
A R T I C LE I N FO
A B S T R A C T
Keywords: ECAP Fracture toughness Fatigue crack growth Crack orientation
In this study, effects of crack plane orientation on the fatigue crack growth rate and fracture toughness of 6063 aluminum alloy deformed by equal channel angular pressing (ECAP) were investigated. The ECAP process continued up to five passes without failure. Grain refinement was obvious after five passes of the ECAP process. Textural studies showed aligning the grains in known directions. After four passes, yield and ultimate strengths increase respectively from 90 MPa and 209 MPa to 300MPa and 375 MPa and also reduction in elongation was observed. The roughness decreased after the process. The fatigue crack growth rate was investigated at different load ranges with the same load ratio for different orientations. The crack growth rate increased after one pass of the ECAP process. After five passes, the AA6063 shows a lower crack growth rate in compared with as-received material. The fracture toughness of mode I and mixed-mode for different orientations were measured. The results showed that the orientation has a significant effect on the fatigue crack growth and fracture toughness of the ECAPed samples. The fracture surfaces were studied using scanning electron microscope (SEM) and refined equiaxed dimples were observed after the ECAP process.
1. Introduction
movement plays an important role in the mechanical properties changes after this process. The fatigue behavior of severely deformed metals was investigated by Vinogradov and Hashimoto [10]. The results showed that the grain boundaries have important effects on both low and high cycle fatigue behaviors of these materials. The SPD materials have shorter low cycle fatigue(LCF) lives in compared with coarse-grained solids because of large decrease in ductility during the process. However, the high cycle fatigue (HCF) lives increased after the SPD process. The shear banding and crack nucleation strongly depends on the material and type of the SPD process. An optimum balance between strength and ductility leads to better fatigue properties [11]. The effects of the ECAP process with subsequent thermal mechanical treatments on the HCF strength of titanium were investigated by Semenova et al. [12]. They found that Low-temperature annealing of samples increased the endurance limit, strength and ductility. Namdar and Jahromi [13] studied the effects of the ECAP on the fatigue behavior of AA2011 alloy. The experimental results revealed that the HCF enhanced by increasing the strength of the material [14]. While the enhancement of tensile strength due to the grain refinement is sensible, the fatigue strength of the UFG materials is not improved in the same manner. The effect of the grain refinement, texture and chemical composition of the material on the fatigue behavior of
The average grain size plays a crucial role in the mechanical properties of a material. Different methods of severe plastic deformations (SPD) were used to produce ultrafine grained (UFG) materials with improved properties such as strength, hardness, fatigue life and fracture toughness. The ECAP method is the common and simple process between the SPD processes. The ECAP process has wide applications for different metallic alloys. This process increased the strength of the material by imposing high strains to introduce the high density of dislocations [1]. The cross-section area of the specimen has no significant changes after the ECAP process. Different routes are available for samples to pass through intersecting channels such as A, B and C routes [2]. The route A is the most widely used route. In this route as can be seen in Fig. 1, the specimen has no rotation according to previous pass. The UFG materials has wide applications in energy (oil-gas), biomedical, aerospace, sports, etc. [3]. Many researchers have studied mechanical properties and microstructures of this materials such as aluminum [4–6] and Mg alloys [7,8]. Horita et al. [9] studied the effects of the ECAP process on the mechanical properties of the aluminum alloys. They found that the strength increased after the ECAP process, however elongation to failure showed a large decrease after the first pass. The dislocation ⁎
Corresponding author. E-mail address: rseifi@basu.ac.ir (R. Seifi).
https://doi.org/10.1016/j.msea.2018.07.042 Received 30 May 2018; Received in revised form 7 July 2018; Accepted 12 July 2018 0921-5093/ © 2018 Elsevier B.V. All rights reserved.
Materials Science & Engineering A 733 (2018) 71–79
M.A. Kazemi, R. Seifi
Fig. 1. Fundamental routes used in ECAP process.
metallic alloys makes it difficult to investigate the influence of only one of these parameters on the fatigue performance. The chemical composition plays crucial role in the fatigue behavior of light alloys [15]. Cavaliere [16] studied the fatigue characteristics of pure metal (Al, Ti, Ni and Cu) produced by the ECAP process. The experimental results showed that the fatigue crack growth rate increased in the pure UFG metals in compared with CG (coarse grained) ones. The fatigue tests revealed that the ECAPed material is less sensitive to the crack initiation due to decreasing the average grain size. Vinogradov et al. [17] investigated the effect of the ECAP process on the fatigue crack growth in copper samples. They found that for small values of ΔKI (stress intensity factor range for mode I), the crack propagation rates for ECAPed copper is higher than that of the CG copper. The growth rate is the same for the UFG and CG materials at higher ΔKI regime. Fatigue crack initiation life for fine grain brass H62 was investigated by Zheng et al. [18]. They found that the finer grains create higher strength in the brass and cause intergranular cracking under the fatigue loadings. Effects of the ECAP process on the fracture toughness of the materials are studied by Seifi and Kazemi [5] for AA6063, Yu et al. [19] for AZ31 and Rahmatabadi et al. [20] for pure AA1050. There are a few studies conducted in field of the effect of the crack plane orientation on the fatigue and fracture behavior of the UFG materials. Hohenwarter and Pippan [21] studied the fracture toughness of the ECAP-deformed iron and effects of the crack plane orientation on the fracture toughness. They examined the fracture toughness in different orientations and measured the critical fracture toughness, KIC and critical J integral, JIC . The results showed that the crack plane orientation has a significant effect on the fracture toughness. Hohenwarter and Pippan [22] investigated the fracture toughness of nanopolycrystalline metals produced by the SPD process. They studied the recent results about the fracture properties of different UFG (d < 1 µm) and nanocrystalline (d < 100 nm) metals. In this study, the effects of the crack plane orientation on the fatigue crack growth of AA6063 aluminum alloy fabricated by the ECAP process have been investigated experimentally. The results were compared with the fatigue and fracture properties of the CG as-received alloy. The microstructure, texture and strength of the ECAPed aluminum were evaluated. The SEM images were taken from fracture surfaces of the samples (ECAPed and CG) to determine the behaviors.
Fig. 2. The cross section of the assembled die.
die was used for the ECAP process as depicted schematically in Fig. 2. The die material was H13 tool steel. This type of the die is suitable for reducing the time of the process, because it doesn’t have bolts. The channels intersecting angle is Φ = 90° and outer corner angle is Ψ = 22°. In order to achieve better grain refinements and mechanical properties, rout A was selected for conducting the ECAP process based on recommendation in [24]. The ECAP process applied to billets up to five passes at 200 °C. The pressing speed of 0.5 mm/s was applied to specimens during the process. Equivalent strain is approximately 1.05 in every pass based on Eq. (1) which presents equivalent strain, εN after N passes of the ECAP process [1]. The strains that imposed on the sample are large simple shear [25].
εN =
N ⎡ Φ+Ψ Φ+Ψ ⎤ ⎞ ⎞ + Ψ cosec ⎛ 2 cot ⎛ 3⎣ ⎝ 2 ⎠⎦ ⎝ 2 ⎠
(1)
The specimens for tensile tests, Compact Tension (CT) and mixedmode samples in different orientations were machined from ECAPed billets. A billet after one pass is shown in Fig. 3. The cross-sectional shape of the billet after extrusion was square. The tensile tests were carried out by using a zwick/roell universal testing machine according to ASTM E8 standard [26]. The tensile specimens were machined from the center of the billets as depicted in Fig. 3(c) with dimensions as follows: gauge length= 45 mm, radius of fillets= 8 mm and diameter= 9 mm. Three tests were performed for each condition. The strain rate and crosshead speed applied on the samples are 3.7 × 10−4s−1 and 1 mm/min, respectively. The microstructure, texture and fracture surface studied by using scanning electron microscope (SEM), atomic force microscope (AFM) and x-ray diffraction (XRD). The standard CT specimens were made from as-received and the ECAPed materials in different orientations according to ASTM E647 [27] for fatigue crack growth and fracture toughness measurements based on ASTM E1820 standard [28]. As depicted in Fig. 4, different types of the CT samples were made and noted as L, M and N. In L type, the crack surface is normal to normal direction (ND) and will grew in extrusion direction (ED), while in M and N types, the crack is normal to the ED. The growth direction of the cracks in M and N types are in the ND and transverse direction (TD), respectively. The width and thickness of the CT specimens were considered as W = 10mm and B = 3mm , respectively. The electric discharge machining (EDM) method was used to create sharp notches in the samples. In order to conduct the fatigue cyclic loading, the servo-hydraulic test
2. Experimental work We used 6063 aluminum alloy (AA6063) for preparing the sample tests. The chemical composition of AA6063 is given in Table 1. The asreceived specimens were annealed at 420 °C for 3 h before ECAP process to achieve better deformability and reduce friction due to the decrease of dislocation density [23]. MoS2 lubricant was used to decrease the friction. Specimens with dimensions of 12 × 12 × 110 mm (Fig. 3a) in form of billets were machined from aluminum rods with 20 mm diameter. A two pieces split Table 1 Chemical composition (wt%) of AA6063. Al
Si
Fe
Cu
Mn
Mg
Cr
Ni
Zn
Ti
Pb
Co
Bi
Base
0.44
0.93
3.50
0.59
1.02
0.03
0.005
0.38
0.04
0.10
0.005
0.004
72
Materials Science & Engineering A 733 (2018) 71–79
M.A. Kazemi, R. Seifi
Fig. 5. a) Schematic view of the modified Arcan fixture and assembled specimens in the fixture b) α = 15° c) α = 60°.
fI = 5.266 − 32.736β + 99.968β 2 − 127.078β 3 + 65.331β 4 fII = 0.3657 + 1.646β − 0.083β 2 Fig. 3. (a) schematic view of As-received Billet, (b) billet after one pass of ECAP, (c) schematic view of tensile sample machined from the center of billet.
3. Results and discussions 3.1. Microstructure and textural studies Fig. 6 shows the microstructure of as-received and ECAPed materials after different passes by using the SEM and AFM techniques. The grain size for as-received aluminum alloy was calculated in the range of 8–75 µm with averaged size as 45 µm. As can be seen, after first pass, the grain size decreased. The crystallite size for the ECAPed materials were calculated using the XRD profile and Williamson-Hall method. This method is valid for the grain size less than 2 µm [32]. The obtained results were confirmed in comparing with work done by Majzoobi et al. [33]. It was assumed that both the crystallite size and the grain size have reduction after the process. The crystallite size after five passes of the ECAP process was measured as 230 nm. The intensity values (in a.u.) for different 2θ degrees are depicted in Fig. 7. The roughness measurements were performed by AFM on the lateral cross-section of as-received and ECAPed specimens. Fig. 8 showed 3D image of the specimens’ surfaces for as-received and five times ECAPed material. The image showed more homogeneous roughness for ECAPed material. The average roughness values were 140 nm (SD=8.4) and 123 nm (SD=7.12) for as-received and ECAPed materials, respectively. Also, pole figures of [2 2 0] and [1 1 1] planes of as-received and ECAPed samples are shown in Fig. 9. The texture of the ECAPed samples depends on the as-received material texture until first pass, and for further passes texture more depends on ECAP process routes. The cubic lattice structures allow slippage to occur more easily than non-cubic lattices, so hcp metals are not as ductile as the fcc metals [34]. There are some deviations textures that departure from the ED, which may be caused by the refinement of the grains. The basal planes are not rotated because of using rout A in the ECAP process. [35]. The results showed that Cθ is the strongest component of texture for specimen after first passes of the ECAP, however, for further passes, Bθ / B′θ is the strongest component. The selected route for ECAP process plays an important role in the texture of the material for further passes of the process [34]. The basal planes in [2 2 0] were oriented parallel to the extrusion direction after five passes. Maximum texture intensity after five passes increased from 5.345 to 10.624 for [2 2 0] plane and from 5.61 to 7.06 for [1 1 1] plane.
Fig. 4. a) CT specimen b) schematic views of L, M and N types.
machine (zwick/roell) was used. The tests were performed at frequency as 30 Hz and load ratio was R = 0.1 for different load ranges. The high resolution camera was used for measuring the crack length at different life times. According to ASTM E647 standard for CT specimens, the stress intensity factor range was calculated by using the following equation:
ΔK =
ΔP f (β ) f (β ) = (2 + β )[0.886 + 4.64β − 13.32β 2 + 14.72β 3 (BBN W )1/2 − 5.6β 4]/(1 − β )3/2
(2)
Where β = a/ W and also ΔP , a, W , B and BN are the load range, crack length, sample width, thickness and net thickness for the samples with side grooves, respectively. If side grooves are not present, then BN = B . For plain strain condition, according to ASTM standard, the sample thickness must be greater than 2.5 mm [29]. The CT and CTS (CT shear) specimens were used for calculation of the fracture toughness under pure modes and mixed-mode loading conditions, respectively. The fatigue pre-crack was created at the end of 4 mm notch. The fracture load-displacement curves were recorded as output of the tests. The ASTM standards E1820 and D5045 [30] were considered. Determination of the mixed-mode fracture toughness has some difficulties due to limitation in the billets dimensions, so Arcan fixture was modified to apply the proper loads on the samples in different angles, as depicted in Fig. 5. The values of mixed-mode stress intensity factors were calculated by employing the following equations [31]:
KI =
P cos α πa Psinα πa f (β ), KII = f (β ), Kt = 1.25BW I 1.25BW II
KI 2 + KII 2
3.2. Tensile strength
(3)
The stress-strain curves for as-received and ECAPed samples are shown in Fig. 10. Young's modulus is about E = 70 GPa and other
KI , KII and Kt are mode I (opening), mode II (sliding mode) and mixed-mode stress intensity factors, respectively. Where: 73
Materials Science & Engineering A 733 (2018) 71–79
M.A. Kazemi, R. Seifi
Fig. 6. SEM images of a) as-received sample (x500), b) as-received (x1800), c) after first pass with x2000, AFM images of TD-ND surfaces d) as-received e) after five passes.
tensile properties given in Table 2. The ECAPed UFG material shows enhancement of dislocations density so the strength would be increased [1]. The remarkable enhancement of tensile and yield strengths occurred after first pass. After the first pass, these values improved by 150% and 45%, respectively while after that until fourth pass (during three passes) improvements are about 233% and 79%, respectively with respect to as-received samples. After the first pass, there are considerable reduction in ductility of the alloy (about 39%), as the results of Horita et al. [9] and after fourth passes, elongation decreased about 58%. 3.3. Fatigue crack growth Under constant range (or amplitude) of the applied stress, based on the relation ΔK = YΔσ πa , stress intensity range ( ΔK ) is proportional with the crack length. For small values of crack length and so small ΔK , the crack propagation is sensitive to grain size. For UFG material, the crack has to break more boundaries because the grain's boundaries are closer to each other [36]. For greater values of ΔK , crack propagation obey a power law manner [37] noted as Paris equation and is less sensitive to grain size.
Fig. 7. Intensity values versus 2θ for a sample after five passes.
74
Materials Science & Engineering A 733 (2018) 71–79
M.A. Kazemi, R. Seifi
Fig. 8. AFM 3D images of the surfaces, a) as-received, b) after five passes.
Fig. 9. Pole figures for as-received and ECAPed AA6063 alloy.
Fig. 11. Fatigue crack length for as-received and ECAPed material at R= 0.1 and Pmax= 200 N.
growth values of as-received and ECAPed materials for M type specimens with R= 0.1 and Pmax= 200 N are shown in Fig. 11. The crack growth rates were increased after first pass and so the life time decreased. The main reason is high decreasing of the elongation and embrittling after first pass [9]. After five passes, AA6063 alloy shows a lower crack growth rate in compared with as-received material. As can be seen, growth rates are low for small cracks in all three cases. Rate increasing is continued up to the different crack lengths. One can see that for as-received sample, crack growth rate increases up to about a= 1.5 mm, while for sample with one pass to about a= 1.0 mm and for five passes to a= 0.75 mm. These behaviors show that with increasing the ECAP passes, the growth rate reaches to its final value in small crack length with more retarding cycles. After these crack lengths, slope of aN curves (growth rates) are almost constant with the same values. Fig. 12 shows the fracture surfaces of the CT specimens for one pass ECAPed material with different sample orientations at R= 0.1 and Pmax = 150 N. There are some minor differences between images. The fracture surfaces of M and N types (Fig. 12(b) and (c)) had the same crack plane orientations. The fracture surface for M and N types specimens indicated more similarity in compared with L type. The fatigue crack growth for one pass ECAPed alloy at different crack plane orientations with R= 0.1 and Pmax= 150 N loading is depicted in Fig. 13. The results show that the crack plane orientation has a significant effect on the crack growth rate. The M and N types specimens have lower rate in compared with L type. The crack growth rate for M type is close to N type because of the same crack plane orientation. The textural study and fracture surfaces in Fig. 12 confirmed the obtained results. Fig. 14 shows the fatigue crack length versus life time for as-received and ECAPed alloy with R= 0.1 at different load ranges. The results reveal that the fatigue life enhanced significantly by decreasing
Fig. 10. Stress-strain curves of AA6063 before and after the ECAP process. Table 2 Tensile properties for as-received and ECAPed samples (all data are the mean values of three tests). sample
Yield Strength (MPa)
Ultimate Strength (MPa)
Elongation (%)
As-received After one pass After four passes
90 225 300
209 304 375
16.6 10.0 7.0
Eq. (4) presents Paris law:
da = C (ΔK )m dN
(4)
where a and N are the crack length and number of the fatigue cycles, respectively. C and m are material parameters [38]. The fatigue crack 75
Materials Science & Engineering A 733 (2018) 71–79
M.A. Kazemi, R. Seifi
Fig. 12. Fracture surfaces for one pass ECAPed samples under Pmax= 150 N and R= 0.1. a) L type, b) M type, c) N type.
Fig. 15. Fatigue crack behavior of Al-6063 after one pass ECAP for different load range at R = 0.1.
maximum load from Pmax= 200 N to Pmax= 150 N. Fig. 11 shows that the maximum effect of the grain refinement under Pmax= 200 N is at beginning of the fatigue crack growth step and for crack length more than 1.5 mm, the three curves of the fatigue crack growth were approximately parallel. As can be seen in Fig. 14 under the load Pmax= 150 N, effect of the grain refinement continued up to a= 1.75 mm. The obtained results reveal that for small load ranges (Pmax=150 N), the effects on the growth rate is increased up to a= 1.75 mm for five passes of the ECAP and up to a= 1.35 mm after one pass. For higher load ranges (Pmax=200 N) effect of the refinement decreased up to a= 0.5 mm for five passes. These results confirmed by other researches [10]. The da/dN versus ΔK at R= 0.1 with different load ranges for the ECAPed alloy is shown in Fig. 15. The fatigue crack behavior of asreceived and five passes ECAPed material with the same load range is depicted in Fig. 16. The fatigue crack growth in Paris regime has significant changes for aluminum alloy before and after ECAP process and is less sensitive to load range. Similar results for effect of the load range on the fatigue crack growth in Paris regime obtained by other researchers [38,39]. The Paris equation parameters, C and m were calculated for
Fig. 13. Fatigue crack length for one pass ECAPed material with different sample orientation at R= 0.1 and max= 150 N.
Fig. 14. Fatigue crack growth for different sample orientations at R= 0.1 and different load ranges. Fig. 16. Fatigue crack behavior of Al-6063 before and after ECAP process at R= 0.1. 76
Materials Science & Engineering A 733 (2018) 71–79
M.A. Kazemi, R. Seifi
Table 3 Paris law parameters for different types of ECAPed AA6063 alloy. ECAPed sample (pass 1)
m
C
L type
5.34
10−12
M type
4.03
10−10
N type
3.98
10−10
different specimen orientations. The results are given in Table 3. As can be seen, the parameters are different for various orientations. The m values increase from 4.03 and 3.98 for M and N types to 5.34 for L type and also C decreases from 10−10 to 10−12 with metric dimensions for rate and SIF values in Paris equation. The M and N types have the same fatigue crack growth behavior while L type sample has higher value.
Fig. 17. Normalized first mode fracture toughness in different passes of different types.
3.4. Fracture toughness
the sample. The loads were applied to the specimens in directions with angles as α = 15° and 60°, where α is the angle between load influence line and notch surface normal vector, as depicted in Fig. 5. The average values of the critical fracture loads were used to determine critical mixed-mode stress intensity factor. The values of unwanted mode III stress intensity factor is negligible due to the perfect symmetry of modified Arcan fixture [40]. A sample load curve and compliance variations lines for mixedmode loading at α = 60° is presented in Fig. 18. These curves were used in determination of the crack driving force which used in Eq. (3) for determining the SIFs in modes I and II. Mixed-mode fracture toughness, Kt was calculated with the above procedure. One of the loading angles (α = 15° ) is close to the opening mode and the other loading angle (α = 60°) is near to the shear loading mode. The fracture behavior of the material in mixed-mode varied independently from mode I for different types, because strength is a function of all types as L, M and N as listed in Table 5. Fig. 19 shows that the normalized fracture toughness, Kt / Kt0 for ECAPed samples and as-received material under various loadings. For α = 15° loading, the results are close to the values of fracture toughness for the pure mode (I) and the same behavior was expected.
The fracture toughness KIC was calculated with mentioned procedure in the ASTM standard (Eq. (2)) and then converted to critical J‐integral, JIC , by employing the following equation [28]:
JIC ≃
KIC 2 (1 − ν 2) E
(5)
The results for KIC and JIC values are presented in Table 4 for asreceived and extruded samples in different passes. After first pass of the ECAP process, the KIC values decreased from 18.4 MPa m for as-received material to 13.1, 15.71 and 14.43 MPa m for L, M and N types, respectively. The major reason of this reduction is decreasing the ductility of the material after the process in comparison with strength improvement. After further passes of the ECAP process, the ductility were decreased with lower rate and strength increased in each passes. Variations of normalized SIFs for mode I versus as-received material (KI / KI0 ) in different types are shown in Fig. 17. KI0 is the fracture toughness of the as-received material. The M and N specimen types showed less reduction of fracture toughness in compare with L type, which was similar to results obtained by Hohenwarter and Pippan [21]. In M and N types, the crack surfaces are normal to ED (Fig. 4) and according to the textural variations, little ductility increasing in these directions is expected. However, in L type, the surfaces are parallel to extrusion and SIFs have more reduction. The results show that crack plane orientation has crucial role in the fracture toughness of the material. For mixed-mode fracture toughness similar to mode I, we considered the crack growth condition as 5% increasing in compliance of
3.5. Fracture surface In order to study the effects of the ECAP process on the fracture surfaces of the samples, SEM images were taken. As can be seen in Fig. 20(a), large and deep dimples due to ductile fracture observed in as-received material. Fig. 20(b) shows the fracture surface of the ECAPed material. As can be seen, the number of dimples increases due to grain refinement after the ECAP process. Fine and shallower dimples are observed in compared with as-received material due to decreasing of the ductility which can be attributed to slow movement of
Table 4 J-integral and fracture toughness values for different passes of the ECAP process and different orientations (each value is mean value of three tests). Pass number As-received 1st
2nd
3rd
4th
5th
type
JIC (kJ / m2)
KIC (MPa m )
L M N L M N L M N L M N L M N
4.38 2.22 3.18 2.69 2.40 3.31 3.00 2.36 3.25 2.83 2.70 3.85 3.59 2.96 4.66 4.17
18.4 13.1 15.7 14.4 13.5 15.9 15.1 13.4 15.7 14.7 14.3 17.1 16.5 15.0 18.8 17.8
Fig. 18. A sample load curve and determination of mixed-mode crack driving force. 77
Materials Science & Engineering A 733 (2018) 71–79
M.A. Kazemi, R. Seifi
Table 5 Fracture toughness and J integral in mixed-mode loading for as-received and ECAPed materials (each value is mean of three tests). Pass number
(α ) degree
Type
KI MPa m
KII MPa m
Kt MPa m
J kJ m−2
As-received
15 60 15
L M N L M N L M N L M N
14.13 7.16 6.28 8.96 7.88 5.83 5.50 6.13 8.86 15.45 9.31 5.89 5.94 6.60
1.71 5.82 0.78 1.12 1.06 4.73 4.47 5.60 1.13 1.94 1.17 4.78 4.82 5.36
14.23 9.22 6.33 9.03 7.95 7.51 7.09 8.31 8.93 15.57 9.39 7.59 7.65 8.51
2.61 1.09 0.51 1.05 0.81 0.73 0.64 0.89 1.03 3.13 1.13 0.74 0.75 0.93
1st pass
60
5th pass
15
60
Fig. 19. mixed-mode fracture toughness for different types under various loadings.
Fig. 20. SEM images of the fracture surfaces: a, d) as-received b) after first passes of the ECAP process c) fracture zones.
4. Conclusion
dislocations in the UFG materials. In Fig. 20(c), two fracture zones: the fatigue crack growth and ductile fracture were specified. Fig. 20(d) shows the fracture surface of as-received material. The fatigue crack deflected along grain boundaries so the fracture surface shows fissured fracture.
Effects of the ECAP process on the fatigue crack growth and fracture behaviors of AA6063 alloy were investigated. Variations of the yield and ultimate strengths, fatigue crack growth rate, pure and mixed mode fracture toughness, texture, roughness, grain refining and fracture 78
Materials Science & Engineering A 733 (2018) 71–79
M.A. Kazemi, R. Seifi
surfaces were analyzed. The following conclusions can be obtained from this study:
[13]
1. Grain refinement was obvious after five passes of the ECAP process. The average grain size reduced from 45μm to about 230nm . 2. The yield and ultimate strength increased significantly after four passes from 90 MPa and 209 MPa to 300MPa and 375 MPa, respectively. The major increasing was after first pass and elongation decreased after four passes from 16.6% to 7%. 3. The roughness decreased for UFG material in compared with CG one. 4. The crack growth rate increases after one pass which is attributed to large decrease of the elongation. After five passes, AA6063 alloy shows a lower crack growth rate in compared with as-received material. The strength after five passes increased while the elongation decreased with small rate up to four passes. 5. The obtained results showed that the crack plane orientation had a significant effect on the crack growth rate and fracture toughness. The M and N types, because of the same crack plane orientation, had lower crack growth rate in compared with L type specimen. The Paris parameters were different for various orientations. 6. The obtained results revealed that for small load ranges, effects of the grain refinement on improvement of fatigue behavior increased and for higher load ranges decreased. 7. The fracture toughness for pure mode I decreased from 18.4 MPa m to 15.71 MPa m for M type samples. The fracture toughness increased after the third pass due to increase in the strength of the material in comparison with as-received one. 8. After the ECAP process fine and shallower dimples were observed in compared with as-received material.
[14]
[15]
[16] [17]
[18] [19]
[20]
[21] [22]
[23]
[24]
[25] [26]
Data availability statement
[27]
The raw/processed data required to reproduce these findings cannot be shared at this time as the data also forms part of an ongoing study.
[28] [29]
References
[30]
[1] R.Z.L. Valiev, G. Terence, Principles of equal-channel angular pressing as a processing tool for grain refinement, Prog. Mater. Sci. 51 (7) (2006) 881–981. [2] M. Gzyl, A. Rosochowski, S. Boczkal, L. Olejnik, The role of microstructure and texture in controlling mechanical properties of AZ31B magnesium alloy processed by I-ECAP, Mater. Sci. Eng.: A 638 (2015) 20–29. [3] Y. Huang, T.G. Langdon, Advances in ultrafine-grained materials, Mater. Today 16 (3) (2013) 85–93. [4] H. Darban, B. Mohammadi, F. Djavanroodi, Effect of equal channel angular pressing on fracture toughness of Al-7075, Eng. Fail. Anal. 65 (2016) 1–10. [5] R. Seifi, M.A. Kazemi, An investigation on the effects of equal channel angular pressing on the mixed‐mode fracture toughness and mechanical properties of 6063 aluminium alloy, Fatigue Fract. Eng. Mater. Struct. (2018), https://doi.org/10. 1111/ffe.12816. [6] E.K. Cardoso, V. Guido, G. Silva, W. Botta Filho, A. Jorge Junior, Microstructural evolution of AA7050 al alloy processed by ECAP, Matéria (Rio de Janeiro) 15 (2010) 291–298. [7] L. Tang, Y. Zhao, R.K. Islamgaliev, C.Y.A. Tsao, R.Z. Valiev, E.J. Lavernia, Y.T. Zhu, Enhanced strength and ductility of AZ80 Mg alloys by spray forming and ECAP, Mater. Sci. Eng.: A 670 (2016) 280–291. [8] L.B. Tong, M.Y. Zheng, X.S. Hu, K. Wu, S.W. Xu, S. Kamado, Y. Kojima, Influence of ECAP routes on microstructure and mechanical properties of Mg–Zn–Ca alloy, Mater. Sci. Eng.: A 527 (16–17) (2010) 4250–4256. [9] Z. Horita, T. Fujinami, M. Nemoto, T.G. Langdon, Improvement of mechanical properties for Al alloys using equal-channel angular pressing, J. Mater. Process. Technol. 117 (3) (2001) 288–292. [10] A. Vinogradov, S. Hashimoto, Fatigue of severely deformed metals, Adv. Eng. Mater. 5 (5) (2003) 351–358. [11] Z.-Z. Hu, Optimum balance between strength and ductility for fatigue properties, Eng. Fract. Mech. 48 (3) (1994) 445–453. [12] I.P. Semenova, R.Z. Valiev, E.B. Yakushina, G.H. Salimgareeva, T.C. Lowe, Strength
[31]
[32]
[33]
[34]
[35]
[36] [37] [38]
[39] [40]
79
and fatigue properties enhancement in ultrafine-grained Ti produced by severe plastic deformation, J. Mater. Sci. 43 (23) (2008) 7354. M. Namdar, S.A.J. Jahromi, Influence of ECAP on the fatigue behavior of agehardenable 2xxx aluminum alloy, Int. J. Miner., Metall., Mater. 22 (3) (2015) 285–291. Y. Hong, C. Sun, The nature and the mechanism of crack initiation and early growth for very-high-cycle fatigue of metallic materials – an overview, Theor. Appl. Fract. Mech. 92 (Supplement C) (2017) 331–350. Y. Estrin, A. Vinogradov, Fatigue behaviour of light alloys with ultrafine grain structure produced by severe plastic deformation: an overview, Int. J. Fatigue 32 (6) (2010) 898–907. P. Cavaliere, Fatigue properties and crack behavior of ultra-fine and nanocrystalline pure metals, Int. J. Fatigue 31 (10) (2009) 1476–1489. A. Vinogradov, T. Kawaguchi, Y. Kaneko, S. Hashimoto, Fatigue crack growth and related microstructure evolution in ultrafine grain copper processed by ECAP, Mater. Trans. 53 (1) (2012) 101–108. M. Zheng, M.X. Tong, H.P. Cai, C.Z. Xu, M.Q. Huang, Fatigue crack initiation life of fine grain brass H62, Theor. Appl. Fract. Mech. 54 (2) (2010) 105–109. X. Yu, Y. Li, L. Li, Fracture mechanism of AZ31 magnesium alloy processed by equal channel angular pressing comparing three point bending test and tensile test, Eng. Fail. Anal. 58 (2015) 322–335. D. Rahmatabadi, R. Hashemi, B. Mohammadi, T. Shojaee, Experimental evaluation of the plane stress fracture toughness for ultra-fine grained aluminum specimens prepared by accumulative roll bonding process, Mater. Sci. Eng.: A 708 (2017) 301–310. A. Hohenwarter, R. Pippan, Fracture of ECAP-deformed iron and the role of extrinsic toughening mechanisms, Acta Mater. 61 (8) (2013) 2973–2983. A. Hohenwarter, R. Pippan, Fracture and fracture toughness of nanopolycrystalline metals produced by severe plastic deformation, Philos. Trans. R. Soc. A: Math., Phys. Eng. Sci. 373 (2038) (2015). G.D. Fan, M.Y. Zheng, X.S. Hu, C. Xu, K. Wu, I.S. Golovin, Effect of heat treatment on internal friction in ECAP processed commercial pure Mg, J. Alloy. Compd. 549 (Supplement C) (2013) 38–45. K.J. Kim, D.Y. Yang, J.W. Yoon, Microstructural evolution and its effect on mechanical properties of commercially pure aluminum deformed by ECAE (Equal Channel Angular Extrusion) via routes A and C, Mater. Sci. Eng.: A 527 (29–30) (2010) 7927–7930. A. Vinogradov, S. Yasuoka, S. Hashimoto, On the effect of deformation mode on fatigue: simple shear vs. pure shear, Mater. Sci. Forum 584–586 (2008) 797–802. A. International, ASTM E8/E8M - 09 Standard Test Methods for Tension Testing of Metallic Materials, ASTM, 2009. A. International, ASTM E647-15e1, Standard Test Method for Measurement of Fatigue Crack Growth Rates, West Conshohocken, 2015. A. International, ASTM E1820-13 Standard Test Method for Measurement of Fracture Toughness, West Conshohocken, 2013. Y. Li, B. Gong, M. Corrado, C. Deng, D. Wang, Experimental investigation of out-ofplane constraint effect on fracture toughness of the SE(T) specimens, Int. J. Mech. Sci. 128–129 (Supplement C) (2017) 644–651. A. International, ASTM D5045-14, Standard Test Methods for Plane-Strain Fracture Toughness and Strain Energy Release Rate of Plastic Materials, West Conshohocken, 2014. N. Hallbäck, N. Jönsson, T-stress evaluations of mixed mode I/II fracture specimens and T-effects on mixed mode failure of aluminium, Int. J. Fract. 76 (2) (1996) 141–168. Y. Zhong, D. Ping, X. Song, F. Yin, Determination of grain size by XRD profile analysis and TEM counting in nano-structured Cu, J. Alloy. Compd. 476 (1–2) (2009) 113–117. G.-H. Majzoobi, J. Nemati, M.K. Pipelzadeh, S. Sulaiman, Characterization of mechanical properties of Al-6063 deformed by ECAE, Int. J. Adv. Manuf. Technol. 84 (1) (2016) 663–672. Mh Shaeri, M. Shaeri, Mt Salehi, S.H. Seyyedein, Mr Abutalebi, Texture evolution of ultrafine grained Al-7075 alloy produced by ECAP, Metall. Eng. 17 (56) (2015) 49–57. D. Hou, T. Liu, D. Shi, H. Chen, H. Chen, Study of twinning behaviors of rolled AZ31 magnesium alloy by interrupted in situ compressive tests, Mater. Sci. Eng.: A 653 (2016) 108–114. M.D. Sangid, H.J. Maier, H. Sehitoglu, The role of grain boundaries on fatigue crack initiation – An energy approach, Int. J. Plast. 27 (5) (2011) 801–821. X. Cai, R. Xia, M. Huo, J. Xu, A threshold formula for fatigue crack growth with mean stress intensity factors, Int. J. Mech. Sci. 135 (2018) 639–645. J. Horky, G. Khatibi, D. Setman, B. Weiss, M.J. Zehetbauer, Effect of microstructural stability on fatigue crack growth behaviour of nanostructured Cu, Mech. Mater. 67 (2013) 38–45. T. Zhao, J. Zhang, Y. Jiang, A study of fatigue crack growth of 7075-T651 aluminum alloy, Int. J. Fatigue 30 (7) (2008) 1169–1180. A.Eh Oskui, N. Choupani, M. Shameli, 3D characterization of mixed-mode fracture toughness of materials using a new loading device, Lat. Am. J. Solids Struct. 13 (2016) 1464–1482.