- Email: [email protected]
Effect of ingot grain refinement on the tensile properties of 2024 Al alloy sheets
Author’s Accepted Manuscript Effect of ingot grain refinement on the tensile properties of 2024 Al alloy sheets Wei Haigen, Xia Fuzhong, Wang. Mingpu
www.elsevier.com/locate/msea
PII: DOI: Reference:
S0921-5093(16)31346-6 http://dx.doi.org/10.1016/j.msea.2016.11.016 MSA34341
To appear in: Materials Science & Engineering A Cite this article as: Wei Haigen, Xia Fuzhong and Wang. Mingpu, Effect of ingot grain refinement on the tensile properties of 2024 Al alloy sheets, Materials Science & Engineering A, http://dx.doi.org/10.1016/j.msea.2016.11.016 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Effect of ingot grain refinement on the tensile properties of 2024 Al alloy sheets Wei Haigen a, Xia Fuzhong a, Wang Mingpu a,* a
School of materials science and engineering, Central South University, Changsha, 410083, China
*
Corresponding author.
E-mail address: [email protected]
Abstract Microstructures of the coarse grain (CG) and fine grain (FG) 2024 Al alloy ingots prepared by different casting procedures are investigated, and the microstructures and tensile properties of the two naturally aged CG and FG 2024 Al alloy sheets prepared from the two ingots are compared. Metallographic observation reveals similar recrystallized grain structures of the two solution treated 2024 Al alloy sheets, but EBSD analysis demonstrates that a non-complete recrystallized substructure is formed in the CG sheet, whereas the FG sheet recrystallizes completely after solution treatment. SEM observation indicates that the remaining phases in the FG sheet are less, smaller and are more dispersedly distributed than that in the CG sheet. These microstructural differences between the two sheets result in slightly lower strengths and higher ductilities in the directions parallel to the rolling plane, and markedly higher strength and ductility in the ND direction for the naturally aged FG sheet. These findings suggest the possibility of improving the mechanical properties, including toughness and fatigue properties, of Al profiles by refining their ingot grain structures.
Keywords: grain refinement; mechanical characterization; aluminum alloys; casting; fracture
Introduction There have been many investigations indicating that the tensile properties, fracture toughness and fatigue properties of Al alloys are determined mainly by their purities, microstructures (including grain size and substructure) and the types, sizes, contents and distributions of the precipitates and impurities in the alloys [1-5]. Based on what has been revealed, several effective ways which can improve the mechanical properties of Al alloys have been proposed and some of 1
them have been put into practice. For example, by reducing the impurity (mainly Fe and Si) content in pure Al, the insoluble particle contents of Al alloys can be reduced consequently and so their fatigue properties are improved [6]. During melting, adding a small amount of (about 0.5 wt.%) Mn, Cr, Sc or Zr into the melt will lead to the formation of a large number of constituent particles in Al alloys which can inhibit the recrystallization during hot working, so that the grain structures of hot rolled or extruded Al profiles can be refined and their mechanical properties are improved [5]. In addition, by adopting a new manufacture technology combining precipitation and hot deformation processings [7], or combining stepped homogenization and hot deformation processings [8], fine grain Al profiles with higher ductility can be produced. Grain refinement, which is commonly used in the aluminum industry in order to improve the casting and hot working capabilities of Al alloy ingots, can be achieved presently by several methods [9-11]. The effect of ingot grain refinement on the ingot mechanical properties has also been extensively studied. It was found that the ingot mechanical properties of 6063 Al alloy [12], 5052 Al alloy [13] and Al-Si alloy [14, 15] can be improved with grain refinement. It was also found that an increased tensile elongation of Al-5Mg alloy ingots at higher temperatures can be achieved by grain refinement which can reduce the solidification cracking susceptibility of the alloy [16]. However, there are few investigations in the literature concerning the effect of ingot grain refinement on the mechanical properties of Al profiles which are manufactured through hot working of Al alloy ingots. Some investigators find that [17], independent of the ingot grain size, after hot working, most of the microstructural features of an Al alloy ingot disappear in its lamella deformation structure which may lead to the conclusion that the ingot grain size has little effect on the microstructure, and hence the mechanical properties of Al profiles. In addition, a certain degree of recrystallization may occur during the hot working and solution treatment of Al profiles, which sometimes results in similar recrystallization grain structures in Al profiles manufactured from the ingots with very different grain sizes, as will be demonstrated in this paper. In this case, according to the principle that material properties depend on its microstructure, the possibility of improving the mechanical properties of Al profiles by refining their ingot grain structures is discarded. In age-hardenable Al alloys, there are always a small amount of impurity particles after the ultimate heat treatment. These insoluble particles can not only reduce the mechanical properties of the alloys [5, 18] but also reduce their corrosion resistance [19]. Ingot refinement, 2
which is reported to be able to improve the distribution homogeneity of second phases in an ingot [20], needs to be studied with respect to its influence on the insoluble particles in Al profiles. In most studies [4, 21, 22], a tensile or fatigue stress in the longitudinal or long-transverse direction of an Al profile is often chosen to evaluate its mechanical properties, whereas its mechanical properties in the short-transverse direction is often hard to be assessed due to the small size of the profile in the direction. It has been revealed [23] that Al profiles always have poor strength and ductility in the short-transverse direction due to that the impurity particles in them are distributed preferentially on the planes perpendicular to their short-transverse direction. So a tension stress in the short-transverse direction will induce the splitting of Al profiles along the preferential distributing planes of the impurity particles. Whether this detrimental mechanical property anisotropy of Al profiles can be improved by refining their ingot grain structures has not yet been studied. In view of the research status outlined above, coarse grain (CG) and fine grain (FG) 2024 Al alloy ingots are prepared, and the microstructures of the two ingots and the hot rolled and solution treated sheets prepared from the two ingots are observed. The mechanical properties of the two naturally aged 2024 Al alloy sheets are compared. Based on what has been revealed, implications of grain refinement of Al alloy ingots are addressed with respect to the mechanical properties of Al profiles.
Materials and methods Pure Al (99.7 wt.% purity), pure Cu, pure Mg and Al-10Mn master alloy were melted in an electric furnace using a graphite crucible to prepare 2024 Al alloy ingots of nominal composition Al-3.95Cu-1.32Mg-0.59Mn (wt.%). In order to obtain the CG and FG 2024 Al alloy ingots, two different procedures were employed. The CG ingot was cast at 730℃, and the FG ingot was cast at 650℃ with 0.1 wt.% Al-5Ti-1B grain refiner added into the melt before casting. After melting both the two ingots were solidified in a graphite mould with wall thickness of 5 mm, bottom thickness of 10 mm and an interior cross section of 100 mm * 100 mm. After solidification, the two ingots were homogenized for 24h at 490℃after which they were hot rolled at 450℃ with a total thickness reduction of 90% to obtain the CG and FG 2024 Al alloy sheets of 10 mm in thickness. During the hot rolling, in order to lighten the edge cracking of the two sheets, they were stress relief annealed at the rolling reductions of 60% and 80% for 15 minutes at 450℃. After the 3
hot rolling, both the CG and FG sheets were solution treated in a salt bath for 1h at 500℃and then naturally aged for one week, after which their tensile properties were measured. In order to fully evaluate the tensile properties, especially in the short-transverse direction, of the two naturally aged 2024 Al alloy sheets, a mini-tensile specimen and a set of matching gripper were designed. The geometry and dimensions of the mini-tensile specimen are shown in Fig. 1a. An image of the gripper made from die steel is shown in Fig. 1b with a fractured tensile specimen. In this study, in order to make the tensile properties of a 2024 Al alloy sheet measured in the RD (0°), at 45°, in the TD (90°) and ND directions comparative, all the tensile specimens of a sheet were prepared in the same shape by wire electron cutting, as shown in Fig. 1c. Tensile tests were conducted using an INSTRON-3382 machine and the average tensile properties from 5 samples were calculated in all cases. One typical tensile curve of the 0°FG 2024 Al alloy specimen is shown in Fig. 1d, from which it can be seen that an initial elongation stage about 1.1% in strain appears before the linear elastic stage of the curve. This is the moving distance of the gripper before it tightens the specimen. It was found in the tests that most of the specimens fractured in the 4 mm long gauge region, as shown in Fig. 1b. The tensile properties in the RD direction of the naturally aged 2024 Al alloy sheet obtained in this study, σ0.2 (yield strength) ≈ 300 MPa，σb (ultimate tensile strength) ≈ 430 MPa，δ (tensile elongation) ≈ 24%, are comparable with the values in the literature [24-26]. From the image of the fractured specimen shown in Fig. 1d it can also be found that its necking mainly occurs within its gauge region. So it was assumed that the part of a tensile curve after the linear elastic stage is the total strain of the 4 mm gauge region of a specimen. As a result, the length of a tensile curve from the post-yield point to the pre-fracture point is measured as the δ of the tensile specimen.
(b) fractured specimen
(a)
4
All dimensions are in mm
(c)
ND 90°
45°
(d)
δ
RD 0° TD
ND necking region
Strain / % Fig. 1. (a) Mini-tensile specimen dimension; (b) the gripper matching the mini-tensile specimen; (c) schematic of different specimen orientations; (d) stress-strain curve of a mini-tensile specimen, to which the necking occurs mainly within its gauge region. During the tensile tests, before the gripper tightens a specimen, an invalid elongation about 1.1% in strain appears on the tensile curve. The length of a tensile curve from the post-yield point to the pre-fracture point is measured as the δ of the tensile specimen. Grain structures of the 2024 Al ingots and sheets were observed under polarized light. The samples were first grinded with a 600# abrasive paper, and then electrolytically polished at -40℃ ~ -30℃ with 20V DC in a solution of HNO3 : CH3OH = 2 : 8, at last anodized with Baker’s reagent (1% HBF4 in H2O) at 5~10℃ with 24V DC. SEM observation, EDS and EBSD analyses were conducted using a FEI Quanta-200 SEM instrument equipped with TSL OIMTM software. Samples for EBSD analysis were prepared according to the preparation procedure of metallographic samples without being anodized as last.
5
Results The grain structures of the CG and FG 2024 Al alloy ingots homogenized for 24h at 490℃ are shown in Fig. 2a and b, respectively, which are similar to the grain structures of the CG and FG ingots before homogenization, respectively. The large difference in grain size between the CG and FG ingots indicates the significant effectiveness of the combined grain refining technology of adding grain refiner and pouring at a low superheat. After hot rolling to the thickness reduction of 90%, the grains in both the two ingots are rolled into a long-pie-like morphology, as shown in Fig. 2c and d which show the grain structures of the hot rolled CG and FG sheets, respectively. After solution treatment for 1h at 500℃, significant recrystallization occurs in both the two sheets, as shown in Fig. 2e and f which show the grain structures of the CG and FG sheets after solution treat, respectively. Recrystallization results in a pan-cake-shaped grain morphology in the two sheets with little difference in grain size between the two sheets. This failure of refining the grain structure of the 2024 Al alloy sheet by refining the ingot grain structure puts forward the necessity of assessing the effect of refining ingot grain structure on other structural characteristics of the 2024 Al alloy sheet.
(c)
(d)
6
(f)
(e)
Fig. 2. Grain structures of (a, b) as-homogenized for 24h at 490℃ (c, d) hot rolled at 450℃ (e, f) solution-treated for 1h at 500℃ samples of (a, c, e) CG and (b, d, f) FG 2024 Al alloys.
Fig. 3a and b show the SEM images of the homogenized CG and FG 2024 Al alloy ingots, respectively, in which the white phases are the second phases remaining in the two ingots after homogenization. No significant difference in the remaining phase content between the two ingots can be found. In addition, EDS analysis reveals that the phase types in the two homogenized ingots are the same, indicating that the ingot grain refinement has no detectable effect on the content and types of remaining phases in the homogenized 2024 Al alloy ingot. Typical common remaining phases in the two ingots after homogenization are the complex AlCuFeMn phases and Al2Cu+Al2CuMg eutectic phases, as shown in Fig. 3c and d, respectively. Nonetheless, because the grains of the CG ingot are so large that some remaining phases in the ingot distributed between parallel dendrite arms beside a dendrite stem form a long phase lath, as shown at location A in Fig. 3a, and some remaining phases within a dendrite grain form parallel phase laths, as shown at location B in Fig. 3a, whereas in the FG ingot, all the remaining phases locate dispersedly on grain boundaries, as shown in Fig. 3b.
A B 7
B
A Fig. 3. SEM images of homogenized (a) CG and (b) FG 2024 Al ingots; (b, c) the common remaining phases in the two homogenized ingots; SEM images of solution treated (e) CG and (f) FG 2024 Al alloy sheets.
Fig. 3e and f show the SEM images of the CG and FG 2024 Al alloy sheets solution treated for 1h at 500℃, respectively. It can be seen that all the remaining phases in both the two sheets are
8
rolled into the particle chains lying parallel to their rolling planes. The first and most obvious difference is the more frequent occurrence of the long particle chains and parallel head-tail particle chains in the CG 2024 Al alloy sheet, as shown at locations A and B in Fig. 3e, respectively, which may be the preserved microstructural characteristics of the CG ingot as that at locations A and B in Fig. 3a, respectively. The second difference is the lower particle content in the FG sheet, even though the two homogenized ingots have approximately the same remaining phase contents, as indicated in Fig. 3a and b. This difference results from the more effective particle breakage occurring in the FG sheet during hot rolling, as will be analyzed later. Fig. 4a and b show the EBSD inverse pole figure (IPF) maps of the solution treated CG and FG 2024 Al alloy sheets, respectively, of which the EBSD image quality (IQ) maps are shown in Fig. 4c and d, respectively, in which the green lines represent the low-angle subgrain boundaries and the blue lines represent the high-angle grain boundaries formed during recrystallization. It is widely accepted that the subgrain boundaries formed during recovery of metals always have a misorientation angle lower than 15°and grain boundaries formed during recrystallization have a misorientation angle higher than 15°[27]. So by marking the subgrain boundaries and grain boundaries in the two sheets, the recrystallization extents in the two sheets can be evaluated. The presence of a large number of subgrains in Fig. 4c indicates that only a partial recrystallization occurs in the CG sheet after solution treatment, whereas the FG 2024 Al alloy sheet recrystallizes completely after solution treatment, as demonstrated in Fig. 4d. The IQ maps of the CG and FG 2024 Al alloy sheets without subgrain boundaries and grain boundaries being marked are shown in Fig, 4e and f, respectively. It can be found that in the CG sheet, those circled regions containing large particles have a completely recrystallized structure, whereas those regions with a recovered substructure have no large particles.
(a)
ND
(b)
RD
9
500μm
500μm
500μm
500μm
500μm Fig. 4. (a, b) EBSD IPF maps and (c, d) corresponding EBSD IQ maps, in which the green and blue lines represent the subgrain boundaries and grain boundaries, respectively, of (a, c) CG and (b, d) FG 2024 Al alloy sheets after solution treatment; (e) and (f) are the same IP maps as (c) and (d), respectively, without boundaries being marked.
10
All the above microstructure observation results indicate that even though the solution treated CG and FG 2024 Al alloy sheets have similar metallographic structures, their substructures and the contents and distributions of the remaining particles in them are different, which may then result in different tensile properties for the naturally aged CG and FG 2024 Al alloy sheets. The tensile properties of the two naturally aged sheets in different directions are shown in Fig. 5. First, neither of two sheets shows in-plane anisotropy, which should be attributed to their recrystallized grain structures, even though only a partial recrystallization has occurred in the CG sheet. Many researches [22, 28-32] concerning the mechanical properties of Al profiles have revealed that an unrecrystallized Al profile often shows an in-plane anisotropy of tensile properties with higher strength and lower ductility in the RD direction than that in the TD (90°) and 45° directions, whereas for the recrystallized Al profiles this in-plane anisotropy vanishes partially or even completely.
σ0.2 σb δ
CG sheet
δ\%
σ0.2, σb \ MPa
FG sheet
0°
45°
90°
ND
Fig. 5. Tensile properties of the naturally aged CG and FG 2024 Al alloy sheets in different directions.
Secondly, the tensile properties of the 0°, 45° and 90° samples of the CG sheet, σ0.2 ≈ 315MPa, σb ≈ 440MPa, δ ≈ 21%, are higher in strength and lower in ductility than that of the FG sheet, σ0.2 ≈ 280MPa, σb ≈ 420MPa, δ ≈ 27%. It is well known that the strength of the naturally aged 2024 Al alloy mainly originates from the age hardening [33]. Obviously in this study the subgrain strengthening mechanism also plays a part in the naturally aged CG sheet as a result of its non-completely recrystallized substructure. The subgrain strengthening mechanism can not only 11
increase an alloy’s strength and reduce its ductility accordingly, but reduce the difference between σ0.2 and σb of the alloy [30]. The difference is approximately 125MPa for the CG 2024 Al alloy sheet, and 140MPa for the FG 2024 Al alloy sheet, indicating that the higher strengths of the 0°, 45° and 90° samples of the CG sheet are caused by the finer substructure of the sheet. It should be noted, however, that even though there are more remaining particles in the CG 2024 Al alloy sheet, as shown in Fig. 3e, the strengths of the 0°, 45° and 90° samples of the CG sheet are not reduced compared with the strengths of the FG sheet in the three directions. This discrepancy has also been noted by other researchers [1, 34, 35]. These authors found that reducing the impurity particle contents in Al alloys by increasing their purities or employing different manufacture technologies cannot increase their strengths in the directions parallel to the preferential distribution planes of the impurity particles in these alloys, such as the rolling plane of a hot rolled Al alloy, but can increase their ductilities in these directions. In view of this, it is concluded here that the higher ductilities of the 0°, 45° and 90° samples of the FG 2024 Al alloy sheet are brought about together by their completely recrystallized substructures and lower impurity contents. At last, the most important result illustrated in Fig. 5 is that the FG 2024 Al alloy sheet has much higher strength and ductility in the ND direction, σ0.2 ≈ 280 MPa, σb ≈ 370 MPa, δ ≈ 7%, than that of the CG sheet, σ0.2 ≈ 260 MPa, σb ≈ 275 MPa, δ ≈ 3%. The differences in tensile properties between the two naturally aged 2024 Al alloy sheets can be further demonstrated from their fracture morphologies. The low magnification SEM images of the fracture surfaces of the CG and FG tensile specimens in the RD direction are illustrated in Fig. 6a and b, respectively, of which the higher magnification images are shown in Fig. 6c and d, respectively. It can be observed that except those regions containing large particles with a cleavage fracture morphology, the remaining fracture surfaces of the two sheets have a ductile fracture morphology with many dimples of 1 μm in size and containing small particles at their bottom. These small particles should be the T phase (Al20Cu2Mn3) in 2024 Al alloy [36]. It is evident that the two 2024 Al alloy sheets have the same facture mechanism, i.e., the ductile fracture mechanism, in the RD direction. It was also found in the experiments that the facture morphologies of the 45° and 90° samples of the two sheets are similar to what are shown in Fig. 6a and b.
12
13
Fig. 6. SEM micrographs of the fracture surfaces of (a, c, e) CG and (b, d, f) FG 2024 Al alloy sheets in the (a, b, c, d) RD and (e, f) ND directions.
The fracture morphologies of the CG and FG sheets in the ND direction are shown in Fig. 6e and f, respectively. It is evident that both the two ND tensile specimens fracture along the remaining particle chains in the two sheets, resulting in the abundance of large particles on their fracture surfaces. It has been observed [37, 38] that the large brittle impurity particles in Al alloys fracture at the early tensile stage, inducing crack sources in a specimen. In this study, as the remaining phases in the homogenized CG and FG ingots are hot rolled into the particle chains lying perpendicular to the ND directions of the two sheets, microcracks along these particle chains can be easily formed during the tensile tests in the ND direction, reducing the actual load-bearing areas in the two ND specimens. A longer particle chain will induce a longer microcrack which can unstably propagate more easily, so the more and longer particle chains present in the CG 2024 Al alloy sheet will dramatically reduce the strength and ductility of the sheet in the ND direction. In addition, it can also be observed from Fig. 6f that more and finer impurity particles are present on the fracture surface of the FG sheet, causing more and finer dimples accordingly than in the CG sheet. This fracture morphological difference between the two sheets evidences the conclusion that more effective particle breakage is caused in the FG sheet during the hot rolling, which is drawn above from the comparison of Fig. 3e and c. It has been reported [2] that the crack-originating capability of an impurity particle in Al alloys decreases dramatically as long as its size is reduced below 6μm, which can be interpreted qualitatively as that a microcrack formed with the fracture of a finer particle has a more weak stress concentration at its tip and so it will be more difficult to propagate. So another important reason for the higher ductility of the FG sheet in the ND direction is the finer remaining particles in it.
(a)
CG sheet FG sheet
14
Elongation \ %
(a)
CG sheet FG sheet
Strain \ %
15
Fig. 7. (a) Tensile stress-strain curves of the CG and FG 2024 Al alloy sheets in the ND direction; side profiles of the corresponding fracture surfaces of (b) CG and (c) FG 2024 Al alloy sheets.
The stress-strain curves of the CG and FG sheets in the ND direction are shown in Fig. 7a, and the side profiles of corresponding fracture surfaces are shown in Fig. 7b and c, respectively. From Fig. 7a it can be seen that the ND specimen of the CG sheet fractures not long after the yield point, without obvious strain hardening before fracture. This should be mainly attributed to the more and longer particle chains in the specimen which can induce more and longer microcracks of which the tip stress concentration is so high that, before significant strain hardening occurs in the metal matrix, the tensile stress has reached the level for the microcracks to unstably propagate. So in the ND specimen of the CG sheet the microcracks propagate much earlier during the tensile tests, reducing the stress abruptly as a result of their cleavage fracture mechanism as reflected in Fig. 6e. In the ND specimen of the FG sheet, the particle chains are less and shorter that not only the microcracks formed along the chains are more difficult to propagate as a result of their lower tip stress concentration, but also these microcracks are more difficult to be connected during propagation to form longer cracks as a result of their larger spacings. So a higher degree of plastic deformation can occur in the ND specimen of the FG sheet to cause a higher level of strain hardening in the metal matrix so as to promote the microcracks to unstably propagate, causing the more ductile fracture morphology of the specimen, as shown in Fig. 6f. It is also those longer particle chains in the CG sheet that lead to the straighter side profile of the fracture surface of the CG specimen in the ND direction, as shown in Fig. 7b, whereas a more curved side profile is formed for the fractured FG specimen, as shown in Fig. 7c, because more frequent crack-direction 16
changes are needed to connect the shorter microcracks formed along the shorter particle chains in the specimen.
Discussion All the above results indicate that the differences in the substructure and the content and distribution of the remaining phases between the two naturally aged sheets lead to slightly lower strengths and higher ductilities of the 0°, 45° and 90° samples and markedly higher strength and ductility of the ND sample for the FG sheet. The microstructural differences between the two 2024 Al alloys sheets not only have something to do with their ingot microstructures which are preserved to some extent in the two sheets after hot rolling, but also are influenced by the different degrees of deformation homogeneities of the two ingots during hot rolling.
(c)
(d)
RD
RD TD
TD
17
Fig. 8. Grain structures of (a) CG and (b) FG 2024 Al alloy sheets hot rolled by a thickness reduction of 60%; IPF maps of the rolling planes of (c) CG and (d) FG 2024 Al alloy sheets hot rolled by a thickness reduction of 90%; (e) and (f) are the corresponding SEM images of (c) and (d), respectively. (g) edge cracking of the two hot rolled sheets at the thickness reduction of 90%.
Microstructures of the CG and FG sheets hot rolled at the thickness reduction of 60% are shown in Fig. 8a and b, respectively. Deformation heterogeneity is evident in the CG sheet, with some severely deformed grains with deformation bands and some slightly deformed grains without obvious metallographic deformation structure, whereas all the grains in the FG sheet are flatly rolled to a similar degree that no obvious deformation heterogeneity can be found in the sheet. With deformation heterogeneity in the CG sheet, the breakage of a grain into different orientation regions during hot rolling is more evident in the sheet, resulting in reduced grain size difference between the hot rolled CG and FG sheets as shown in Fig. 2c and d, respectively, compared to the grain size difference between the CG and FG ingots shown in Fig. 2a and b, respectively. Fig. 8c and d show the IPF maps on the rolling planes of CG and FG sheets at the thickness reduction of 90%, respectively. Fig. 8e and f show the corresponding SEM images in
18
which the circled holes are where the large remaining particles locate which are preferentially polished during the sample preparation. Other smaller holes on the sample surfaces are the defects caused after the precipitated particles (Al2CuMg) formed during the hot rolling are preferentially polished. It can be found in Fig. 8c and d that significant recrystallization occurs in the two sheets during the hot rolling around the large particles. Particles in Al alloys above a critical size are effective to stimulate recrystallization during the hot working [39]. The recrystallization grains in the CG sheet are larger than that in the FG sheet, as can be seen in Fig. 8c and d. Recrystallization during hot rolling can cause recrystallization softening, inducing strain concentration in these recrystallized regions. As a result, the strain concentration should be higher in the CG sheet as a result of the larger recrystallization grains in the sheet, and so the unrecrystallized regions in the CG sheet will experience lower deformation than that in the FG sheet. This conclusion is supported by the result that higher orientation variation is present in the unrecrystallized regions in the FG sheet, as shown in Fig. 8c and d, suggesting higher degree of deformation heterogeneity in the CG sheet. The edge cracking of the two sheets at the thickness reduction of 90% are shown in Fig. e. It is evident that the cracks on the CG sheet edge are longer, deeper and are more unevenly distributed. The higher susceptibility of the CG sheet to the edge cracking has a relationship with the higher strain concentration in the recrystallized regions of the sheet, which was also demonstrated by others [40]. The higher degree of deformation heterogeneity also leads to a higher degree of stored energy heterogeneity in the CG sheet after hot rolling, which then leads to the non-complete recrystallization substructure in the sheet after solution treatment, as shown in Fig. 4e. This suggests that the recrystallization in the two sheets is affected not only by the large particles but also by their deformation homogeneities during hot rolling. According to some investigations [22, 23], increasing the hot working temperature of Al profiles will increase their recrystallization temperature accordingly. This is because a higher hot working temperature will results in a higher degree of dynamic recovery during the hot working, reducing the stored deformation energy in Al profiles and so increasing their recrystallization temperature. So it is suggested here that by controlling the hot working temperature of an FG Al alloy ingot, the substructure of its profile after the ultimate heat treatment can be chosen purposely to be the completely recrystallized or recovered one, which is useful for the application requirement that the deformation structure of an 19
Al profile needs to be preserved after the ultimate heat treatment so as to preserve its higher strength in the rolling or extrusion direction. In this case, adapting an Al profile manufactured from the FG ingots with recovered substructure can not only prevent the recrystallization softening, but endow the profile with higher ductility as a result of the reduced size and improved distribution of the remaining large particles in the Al profile. For an Al profile manufactured from a CG Al ingot it is more difficult to control its substructure by controlling the hot working temperature as a result of its heterogeneous deformation substructure. For the on-line quenching Al profiles, a finer ingot grain structure will bring about a finer microstructure for the Al profiles, as shown in Fig. 2d, which is evidently beneficial for improving their mechanical properties. It can be observed from Fig. 3a and b that some of the remaining phases in the CG ingot distribute between dendrite arms, i.e., within the grains, whereas all the remaining phases in the FG ingot locate at grain boundaries. It is well known that during the deformation of metals the regions near grain boundaries experience more severe deformation than that within grains [41]. This will result in a more effective breakage of the remaining phases in the FG ingot. In addition, because some grains in the CG ingot experience a slightly weaker deformation, as shown in Fig. 8a and c, the remaining particles in these grains will be slightly broken. Both the two above conditions will result in the less effective particle breakage in the CG 2024 Al alloy sheet, as illustrated in Figs. 3e and 6e. The improved strength and ductility of the FG 2024 Al alloy sheet in the ND direction will endow the sheet with higher service reliability. It is shown in Fig. 9a the schematic diagram of the microstructure at the base location of an I type extruded Al profile [30]. First, a stress concentration is more likely to occur at this corner location. Second, the fibrous grain structure at P location leads the profile to be vulnerable to the splitting along grain boundaries containing impurity particles under a tensile stress either in the vertical or in the horizontal direction. In this case, it is particularly beneficial to use a FG Al alloy ingot to manufacture the Al profile so as to increase its strengths and ductilities in the vertical and horizontal directions.
20
(a) p
RD and Tensile axis
Fig. 9. (a) Schematic diagram of the fibrous grain structure at the base location of an I type extruded Al profile [30]; (b) side profile of the fracture surface of the CG 2024 Al alloy specimen in the RD direction. The shear failure surface of the specimen is disconnected by the splitting (indicated by red lines) along the remaining particle chains.
Fig. 9b shows the side profile of the fracture surface of the CG 2024 Al alloy specimen in the RD direction. It can be seen that even though its shear failure surfaces are inclined at 45° to the tensile direction, those microcracks in the RD direction (indicated by red lines), which should be caused by the splitting along the particle chains in the specimen, disconnect the main crack onto different shear planes. It was reported that [42], even without reducing the strength of an Al profile, this influence of impurity particles on the crack propagation can reduce the toughness of the Al profile by rendering a Zig-Zag or twisted-type of crack propagation pattern. In addition, reducing the size and number of the impurity particles in Al profiles is beneficial for improving their fatigue properties [18, 43]. In view of this, FG Al alloy ingots are suggested to be used so as to improve the mechanical properties of Al profiles, because in current industrial practices, it is common to adopt Al alloy ingots with dendrite grains of several hundreds of μm in diameter to produce Al profiles.
21
Conclusions By employing a combined casting technology of adding 0.1 wt.% Al-5Ti-1B grain refiner and pouring at a low superheat (650℃), solidification structure of the 2024 Al alloy ingot is significantly refined. However, this grain refinement has little effect on the types and content of the remaining phases in the ingot after homogenization, but leads to a more even distribution of the remaining phases. The grain refinement also leads to a more homogenous deformation and lower degree of recrystallization in the FG sheet during hot rolling as a result of the finer recrystallization grains in the sheet, and hence lightens the edge-cracking of the sheet. Even though the grain sizes of the two solution treated 2024 Al alloy sheets are the same, a non-complete and complete recrystallization occur in the CG and FG sheets, respectively. With more dispersedly distributed remaining phases in the homogenized FG ingot and more effective breakage of the remaining particles during the hot rolling in the FG sheet. the remaining particles are smaller and the remaining particle chains are shorter and less in the FG sheet. These above microstructural differences between the two sheets result in slightly lower strengths and higher ductilities of the 0°, 45° and 90° samples and markedly higher strength and ductility of the ND sample for the FG sheet, which suggests that the benefits of ingot grain refinement should not be narrowly regarded as being able to improve the casting and hot working capabilities of an ingot.
References [1] J.S. Santner, A study of fracture in high purity 7075 aluminum alloys, Metall. Trans. A 9 (1978) 769-779. http://dx.doi.org/10.1007/BF02649786 [2] C.Y. Kung, M.E. Fine, Fatigue crack initiation and microcrack growth in 2024-T4 and 2124-T4 aluminum alloys, Metall. Trans. A 10 (1979) 603-610. http://dx.doi.org/doi:10.1007/BF02658324 [3] N.E. Prasad, S.V. Kamat, K.S. Prasad, G. Malakondaiah, V.V. Kutumbarao, In-plane anisotropy in the fracture toughness of an Al-Li 8090 alloy plate, Eng. Fract. Mech. 46 (1993) 209-223. http://dx.doi.org/10.1016/0013-7944(93)90282-W [4] The effect of grain size on low-cycle fatigue behavior of Al-2024 polycrystalline alloy, Metall. Mater. Trans. A 35 (2004) 2725-2728. http://dx.doi.org/10.1007/s11661-004-0218-1 [5] T. Warner, Recently-developed aluminium solutions for aerospace applications, Mater. Sci. Forum. 519 (2006) 1271-1278. http://dx.doi.org/10.4028/www.scientific.net/MSF.519-521.127 [6] J. Staley, Influence of microstructure on fatigue and fracture of aluminum alloys, Aluminum 55 (1979) 277-281. 22
[7] N.E. Paton, C.H. Hamilton, J.A. Wert, M.W. Mahoney, Characterization of fine-grained superplastic aluminum alloys, JOM 34 (1982) 21-27. http://dx.doi.org/doi:10.1007/BF03338068 [8] J. Waldman, H. Sulinski, H. Markus, The effect of ingot processing treatments on the grain size and properties of Al alloy 7075, Metall. Trans. 5 (1974) 573-584. http://dx.doi.org/10.1007/bf02644652 [9] B.S. Murty, S.A. Kori, M. Chakraborty, Grain refinement of aluminium and its alloys by heterogeneous
nucleation
and
alloying,
Int.
Mater.
Rev.
47
(2002)
3-29.
http://dx.doi.org/10.1179/095066001225001049 [10] T. Campanella, C. Charbon, M. Rappaz, Grain refinement induced by electromagnetic stirring: A dendrite
fragmentation
criterion,
Metall.
Mater.
Trans.
A
35
(2004)
3201-3210.
http://dx.doi.org/10.1007/s11661-004-0064-1 [11] M.A. Easton, H. Kaufmann, W. Fragner, The effect of chemical grain refinement and low superheat pouring on the structure of NRC castings of aluminium alloy Al–7Si–0.4Mg, Mater. Sci. Eng. A 420 (2006) 135-143. http://dx.doi.org/10.1016/j.msea.2006.01.078 [12] E. Wang, T. Gao, J. Nie, X. Liu, Grain refinement limit and mechanical properties of 6063 alloy inoculated
by
Al–Ti–C
(B)
master
alloys,
J.
Alloy.
Compd.
594
(2014)
7-11.
http://dx.doi.org/10.1016/j.jallcom.2014.01.145 [13] A.B. Pattnaik, S. Das, B.B. Jha, N. Prasanth, Effect of Al–5Ti–1B grain refiner on the microstructure, mechanical properties and acoustic emission characteristics of Al5052 aluminium alloy, J. Mater. Res. Technol. 4 (2015) 171-179. http://dx.doi.org/10.1016/j.jmrt.2014.10.017 [14] S. Meenia, F. Khan Md, S. Babu, R.J. Immanuel, S.K. Panigrahi, G.D. Janaki Ram, Particle refinement and fine-grain formation leading to enhanced mechanical behaviour in a hypo-eutectic Al–Si alloy subjected to multi-pass friction stir processing, Mater. Charact. 113 (2016) 134-143. http://dx.doi.org/10.1016/j.matchar.2016.01.011 [15] L. Bolzoni, M. Nowak, N. Hari Babu, Grain refinement of Al–Si alloys by Nb–B inoculation. Part II: Application
to
commercial
alloys,
Mater.
Design.
66
(2015)
376-383.
http://dx.doi.org/10.1016/j.matdes.2014.08.067 [16] R. Takai, S. Kimura, R. Kashiuchi, H. Kotaki, M. Yoshida, Grain refinement effects on the strain rate sensitivity and grain boundary sliding in partially solidified Al–5 wt%Mg alloy, Mater. Sci. Eng. A 667 (2016) 417-425. http://dx.doi.org/10.1016/j.msea.2016.05.023 [17] P.N. Anyalebechi, Characterization of the Microstructure of Commercial-Size Ingots of Aluminum Alloy
3004,
EPD
Congress
2011,
John
Wiley
&
Sons,
Inc.,
pp.
678-700.
http://dx.doi.org/10.1002/9781118495285.ch75 [18] J.H. Mulherin, H. Rosenthal, Influence of nonequilibrium second-phase particles formed during solidification upon the mechanical behavior of an aluminum alloy, Metall. Trans. 2 (1971) 427-432. http://dx.doi.org/10.1007/bf02663329 [19] W. Zhang, G.S. Frankel, Transitions between pitting and intergranular corrosion in AA2024, Electrochim. Acta. 48 (2003) 1193-1210. http://dx.doi.org/10.1016/S0013-4686(02)00828-9 [20] W. Shuncheng, G. Chunlei, L. Xiaohui, Z. Kaihong, Q. Wenjun, Microstructure of Al-4.99Zr-1.1B Master Alloy and Its Grain Refinement Effect on AZ31 Magnesium Alloy, Rare. Metal. Mat. Eng. 43 (2014) 2567-2571. http://dx.doi.org/10.1016/S1875-5372(15)60002-9 [21] J. Mizera, J.H. Driver, E. Jezierska, K.J. Kurzydłowski, Studies of the relationship between the microstructure and anisotropy of the plastic properties of industrial aluminum-lithium alloys, Mater. Sci. Eng. A 212 (1996) 94-101. http://dx.doi.org/10.1016/0921-5093(96)10172-6 [22] K.V. Jata, S. Panchanadeeswaran, A.K. Vasudevan, Evolution of texture, micro structure and
23
mechanical property anisotropy in an Al-Li-Cu alloy, Mater. Sci. Eng. A 257 (1998) 37-46. http://dx.doi.org/10.1016/S0921-5093(98)00822-3 [23] K. TAKAHASHI, K. MINAKAWA, C. OUCHI, The effect of thermomechanical processing variables on anisotropy in mechanical properties of Al-Li alloys, J. Phys. Colloques. 48 (1987) C3-163-C163-169. http://dx.doi.org/10.1051/jphyscol:1987319 [24] A. Heinz, A. Haszler, C. Keidel, S. Moldenhauer, R. Benedictus, W.S. Miller, Recent development in aluminium alloys for aerospace applications, Mater. Sci. Eng. A 280 (2000) 102-107. http://dx.doi.org/10.1016/S0921-5093(99)00674-7 [25] G. Bussu, P.E. Irving, The role of residual stress and heat affected zone properties on fatigue crack propagation in friction stir welded 2024-T351 aluminium joints, Int. J. Fatigue. 25 (2003) 77-88. http://dx.doi.org/10.1016/S0142-1123(02)00038-5 [26] R. Zettler, S. Lomolino, J.F. dos Santos, T. Donath, F. Beckmann, T. Lippman, D. Lohwasser, Effect of Tool Geometry and Process Parameters on Material Flow in FSW of an AA 2024-T351 Alloy, Weld World 49 (2005) 41-46. http://dx.doi.org/10.1007/bf03266474 [27] S. Gourdet, F. Montheillet, An experimental study of the recrystallization mechanism during hot deformation
of
aluminium,
Mater.
Sci.
Eng.
A
283
(2000)
274-288.
http://dx.doi.org/10.1016/S0921-5093(00)00733-4 [28] R.E. Zinkham, Anisotropy and thickness effects in fracture of 7075-T6 and -T651 aluminum alloy, Eng. Fract. Mech. 1 (1968) 275-276. http://dx.doi.org/10.1016/0013-7944(68)90003-9 [29] M.H. SKILLINGBERG, R.F. ASHTON, Processing and performance of Al-Li-Cu-X extrusions, J. Phys. Colloques 48 (1987) C3-179-C173-186. http://dx.doi.org/10.1051/jphyscol:1987321 [30] S.J. Hales, R.A. Hafley, Texture and anisotropy in Al-Li alloy 2195 plate and near-net-shape extrusions, Mater. Sci. Eng. A 257 (1998) 153-164. http://dx.doi.org/10.1016/S0921-5093(98)00834-X [31] R.K. Singh, A.K. Singh, N.E. Prasad, Texture and mechanical property anisotropy in an Al–Mg–Si– Cu alloy, Mater. Sci. Eng. A 277 (2000) 114-122. http://dx.doi.org/10.1016/S0921-5093(99)00549-3 [32] W. Hasso, Industrial application of recrystallization control in aluminum products, Mater. Sci. Forum. 467 (2004) 349-356. http://dx.doi.org/10.4028/www.scientific.net/MSF.467-470.349 [33] R.K.W. Marceau, G. Sha, R. Ferragut, A. Dupasquier, S.P. Ringer, Solute clustering in Al–Cu–Mg alloys during the early stages of elevated temperature ageing, Acta. Mater. 58 (2010) 4923-4939. http://dx.doi.org/10.1016/j.actamat.2010.05.020 [34] P.J. Blau, Influence of Iron and Silicon Content on the Tensile Properties of 7X75 and ZR-Modified 7X75 Aluminum Plate, Air Force Materials Lab., Wright-Patterson AFB, Ohio (USA). [35] D.S. Thompson, Metallurgical factors affecting high strength aluminum alloy production, Metall. Trans. A 6 (1975) 671-683. http://dx.doi.org/10.1007/BF02672287 [36] Y.Q. Chen, D.Q. Yi, Y. Jiang, B. Wang, D.Z. Xu, S.C. Li, Twinning and orientation relationships of T-phase
precipitates
in
an
Al
matrix,
J.
Mater.
Sci.
48
(2013)
3225-3231.
http://dx.doi.org/10.1007/s10853-012-7102-y [37] G.T. Hahn, A.R. Rosenfield, Metallurgical factors affecting fracture toughness of aluminum alloys, Metall. Trans. A 6 (1975) 653-668. http://dx.doi.org/10.1007/bf02672285 [38] R.H. Van Stone, J.A. Psioda, Discussion of “metallurgical factors affecting fracture toughness of aluminum alloys”, Metall. Trans. A 6 (1975) 668-670. http://dx.doi.org/10.1007/bf02672286 [39] J.D. Robson, Microstructural evolution in aluminium alloy 7050 during processing, Mater. Sci. Eng. A 382 (2004) 112-121. http://dx.doi.org/10.1016/j.msea.2004.05.006 [40] J.J. Sundelin, S.T. Nurmi, T.K. Lepistö, Recrystallization behaviour of SnAgCu solder joints, Mater.
24
Sci. Eng. A 474 (2008) 201-207. http://dx.doi.org/10.1016/j.msea.2007.04.013 [41] W.Z. Abuzaid, M.D. Sangid, J.D. Carroll, H. Sehitoglu, J. Lambros, Slip transfer and plastic strain accumulation across grain boundaries in Hastelloy X, J Mech. Phys. Solids 60 (2012) 1201-1220. http://dx.doi.org/10.1016/j.jmps.2012.02.001 [42] H.X. Li, C.Q. Chen, Mechanism of anisotropy in fracture behaviour and fracture toughness of high strength
aluminium
alloy
plate,
Mater.
Sci.
Tech-lond.
6
(1990)
850-856.
http://dx.doi.org/10.1179/mst.1990.6.9.850 [43] J. Grosskreutz, G. Shaw, D. Benson, The effect of inclusion size and distribution on fatigue of 2024-T4 aluminum, Wright-Patterson Air Force Base, Dayton, Ohio, USA.
25
Figure captions Fig. 1. (a) Mini-tensile specimen dimension; (b) the gripper matching the mini-tensile specimen; (c) schematic of different specimen orientations; (d) stress-strain curve of a mini-tensile specimen, to which the necking occurs mainly within its gauge region. During the tensile tests, before the gripper tightens a specimen, an invalid elongation about 1.1% in strain appears on the tensile curve. The length of a tensile curve from the post-yield point to the pre-fracture point is measured as the δ of the tensile specimen. Fig. 2. Grain structures of (a, b) as-homogenized for 24h at 490℃ (c, d) hot rolled at 450℃ (e, f) solution-treated for 1h at 500℃ samples of (a, c, e) CG and (b, d, f) FG 2024 Al alloys. Fig. 3. SEM images of homogenized (a) CG and (b) FG 2024 Al ingots; (b, c) the common remaining phases in the two homogenized ingots; SEM images of solution treated (e) CG and (f) FG 2024 Al alloy sheets. Fig. 4. (a, b) EBSD IPF maps and (c, d) corresponding EBSD IQ maps, in which the green and blue lines represent the subgrain boundaries and grain boundaries, respectively, of (a, c) CG and (b, d) FG 2024 Al alloy sheets after solution treatment; (e) and (f) are the same IP maps as (c) and (d), respectively, without boundaries being marked. Fig. 5. Tensile properties of the naturally aged CG and FG 2024 Al alloy sheets in different directions. Fig. 6. SEM micrographs of the fracture surfaces of (a, c, e) CG and (b, d, f) FG 2024 Al alloy sheets in the (a, b, c, d) RD and (e, f) ND directions. Fig. 7. (a) Tensile stress-strain curves of the CG and FG 2024 Al alloy sheets in the ND direction; side profiles of the corresponding fracture surfaces of (b) CG and (c) FG 2024 Al alloy sheets. Fig. 8. Grain structures of (a) CG and (b) FG 2024 Al alloy sheets hot rolled by a thickness reduction of 60%; IPF maps of the rolling planes of (c) CG and (d) FG 2024 Al alloy sheets hot rolled by a thickness reduction of 90%; (e) and (f) are the corresponding SEM images of (c) and (d), respectively. (g) edge cracking of the two hot rolled sheets at the thickness reduction of 90%. Fig. 9. (a) Schematic diagram of the fibrous grain structure at the base location of an I type extruded Al profile [30]; (b) side profile of the fracture surface of the CG 2024 Al alloy specimen in the RD direction. The shear failure surface of the specimen is disconnected by the splitting (indicated by red lines) along the remaining particle chains.
26
www.elsevier.com/locate/msea
PII: DOI: Reference:
S0921-5093(16)31346-6 http://dx.doi.org/10.1016/j.msea.2016.11.016 MSA34341
To appear in: Materials Science & Engineering A Cite this article as: Wei Haigen, Xia Fuzhong and Wang. Mingpu, Effect of ingot grain refinement on the tensile properties of 2024 Al alloy sheets, Materials Science & Engineering A, http://dx.doi.org/10.1016/j.msea.2016.11.016 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Effect of ingot grain refinement on the tensile properties of 2024 Al alloy sheets Wei Haigen a, Xia Fuzhong a, Wang Mingpu a,* a
School of materials science and engineering, Central South University, Changsha, 410083, China
*
Corresponding author.
E-mail address: [email protected]
Abstract Microstructures of the coarse grain (CG) and fine grain (FG) 2024 Al alloy ingots prepared by different casting procedures are investigated, and the microstructures and tensile properties of the two naturally aged CG and FG 2024 Al alloy sheets prepared from the two ingots are compared. Metallographic observation reveals similar recrystallized grain structures of the two solution treated 2024 Al alloy sheets, but EBSD analysis demonstrates that a non-complete recrystallized substructure is formed in the CG sheet, whereas the FG sheet recrystallizes completely after solution treatment. SEM observation indicates that the remaining phases in the FG sheet are less, smaller and are more dispersedly distributed than that in the CG sheet. These microstructural differences between the two sheets result in slightly lower strengths and higher ductilities in the directions parallel to the rolling plane, and markedly higher strength and ductility in the ND direction for the naturally aged FG sheet. These findings suggest the possibility of improving the mechanical properties, including toughness and fatigue properties, of Al profiles by refining their ingot grain structures.
Keywords: grain refinement; mechanical characterization; aluminum alloys; casting; fracture
Introduction There have been many investigations indicating that the tensile properties, fracture toughness and fatigue properties of Al alloys are determined mainly by their purities, microstructures (including grain size and substructure) and the types, sizes, contents and distributions of the precipitates and impurities in the alloys [1-5]. Based on what has been revealed, several effective ways which can improve the mechanical properties of Al alloys have been proposed and some of 1
them have been put into practice. For example, by reducing the impurity (mainly Fe and Si) content in pure Al, the insoluble particle contents of Al alloys can be reduced consequently and so their fatigue properties are improved [6]. During melting, adding a small amount of (about 0.5 wt.%) Mn, Cr, Sc or Zr into the melt will lead to the formation of a large number of constituent particles in Al alloys which can inhibit the recrystallization during hot working, so that the grain structures of hot rolled or extruded Al profiles can be refined and their mechanical properties are improved [5]. In addition, by adopting a new manufacture technology combining precipitation and hot deformation processings [7], or combining stepped homogenization and hot deformation processings [8], fine grain Al profiles with higher ductility can be produced. Grain refinement, which is commonly used in the aluminum industry in order to improve the casting and hot working capabilities of Al alloy ingots, can be achieved presently by several methods [9-11]. The effect of ingot grain refinement on the ingot mechanical properties has also been extensively studied. It was found that the ingot mechanical properties of 6063 Al alloy [12], 5052 Al alloy [13] and Al-Si alloy [14, 15] can be improved with grain refinement. It was also found that an increased tensile elongation of Al-5Mg alloy ingots at higher temperatures can be achieved by grain refinement which can reduce the solidification cracking susceptibility of the alloy [16]. However, there are few investigations in the literature concerning the effect of ingot grain refinement on the mechanical properties of Al profiles which are manufactured through hot working of Al alloy ingots. Some investigators find that [17], independent of the ingot grain size, after hot working, most of the microstructural features of an Al alloy ingot disappear in its lamella deformation structure which may lead to the conclusion that the ingot grain size has little effect on the microstructure, and hence the mechanical properties of Al profiles. In addition, a certain degree of recrystallization may occur during the hot working and solution treatment of Al profiles, which sometimes results in similar recrystallization grain structures in Al profiles manufactured from the ingots with very different grain sizes, as will be demonstrated in this paper. In this case, according to the principle that material properties depend on its microstructure, the possibility of improving the mechanical properties of Al profiles by refining their ingot grain structures is discarded. In age-hardenable Al alloys, there are always a small amount of impurity particles after the ultimate heat treatment. These insoluble particles can not only reduce the mechanical properties of the alloys [5, 18] but also reduce their corrosion resistance [19]. Ingot refinement, 2
which is reported to be able to improve the distribution homogeneity of second phases in an ingot [20], needs to be studied with respect to its influence on the insoluble particles in Al profiles. In most studies [4, 21, 22], a tensile or fatigue stress in the longitudinal or long-transverse direction of an Al profile is often chosen to evaluate its mechanical properties, whereas its mechanical properties in the short-transverse direction is often hard to be assessed due to the small size of the profile in the direction. It has been revealed [23] that Al profiles always have poor strength and ductility in the short-transverse direction due to that the impurity particles in them are distributed preferentially on the planes perpendicular to their short-transverse direction. So a tension stress in the short-transverse direction will induce the splitting of Al profiles along the preferential distributing planes of the impurity particles. Whether this detrimental mechanical property anisotropy of Al profiles can be improved by refining their ingot grain structures has not yet been studied. In view of the research status outlined above, coarse grain (CG) and fine grain (FG) 2024 Al alloy ingots are prepared, and the microstructures of the two ingots and the hot rolled and solution treated sheets prepared from the two ingots are observed. The mechanical properties of the two naturally aged 2024 Al alloy sheets are compared. Based on what has been revealed, implications of grain refinement of Al alloy ingots are addressed with respect to the mechanical properties of Al profiles.
Materials and methods Pure Al (99.7 wt.% purity), pure Cu, pure Mg and Al-10Mn master alloy were melted in an electric furnace using a graphite crucible to prepare 2024 Al alloy ingots of nominal composition Al-3.95Cu-1.32Mg-0.59Mn (wt.%). In order to obtain the CG and FG 2024 Al alloy ingots, two different procedures were employed. The CG ingot was cast at 730℃, and the FG ingot was cast at 650℃ with 0.1 wt.% Al-5Ti-1B grain refiner added into the melt before casting. After melting both the two ingots were solidified in a graphite mould with wall thickness of 5 mm, bottom thickness of 10 mm and an interior cross section of 100 mm * 100 mm. After solidification, the two ingots were homogenized for 24h at 490℃after which they were hot rolled at 450℃ with a total thickness reduction of 90% to obtain the CG and FG 2024 Al alloy sheets of 10 mm in thickness. During the hot rolling, in order to lighten the edge cracking of the two sheets, they were stress relief annealed at the rolling reductions of 60% and 80% for 15 minutes at 450℃. After the 3
hot rolling, both the CG and FG sheets were solution treated in a salt bath for 1h at 500℃and then naturally aged for one week, after which their tensile properties were measured. In order to fully evaluate the tensile properties, especially in the short-transverse direction, of the two naturally aged 2024 Al alloy sheets, a mini-tensile specimen and a set of matching gripper were designed. The geometry and dimensions of the mini-tensile specimen are shown in Fig. 1a. An image of the gripper made from die steel is shown in Fig. 1b with a fractured tensile specimen. In this study, in order to make the tensile properties of a 2024 Al alloy sheet measured in the RD (0°), at 45°, in the TD (90°) and ND directions comparative, all the tensile specimens of a sheet were prepared in the same shape by wire electron cutting, as shown in Fig. 1c. Tensile tests were conducted using an INSTRON-3382 machine and the average tensile properties from 5 samples were calculated in all cases. One typical tensile curve of the 0°FG 2024 Al alloy specimen is shown in Fig. 1d, from which it can be seen that an initial elongation stage about 1.1% in strain appears before the linear elastic stage of the curve. This is the moving distance of the gripper before it tightens the specimen. It was found in the tests that most of the specimens fractured in the 4 mm long gauge region, as shown in Fig. 1b. The tensile properties in the RD direction of the naturally aged 2024 Al alloy sheet obtained in this study, σ0.2 (yield strength) ≈ 300 MPa，σb (ultimate tensile strength) ≈ 430 MPa，δ (tensile elongation) ≈ 24%, are comparable with the values in the literature [24-26]. From the image of the fractured specimen shown in Fig. 1d it can also be found that its necking mainly occurs within its gauge region. So it was assumed that the part of a tensile curve after the linear elastic stage is the total strain of the 4 mm gauge region of a specimen. As a result, the length of a tensile curve from the post-yield point to the pre-fracture point is measured as the δ of the tensile specimen.
(b) fractured specimen
(a)
4
All dimensions are in mm
(c)
ND 90°
45°
(d)
δ
RD 0° TD
ND necking region
Strain / % Fig. 1. (a) Mini-tensile specimen dimension; (b) the gripper matching the mini-tensile specimen; (c) schematic of different specimen orientations; (d) stress-strain curve of a mini-tensile specimen, to which the necking occurs mainly within its gauge region. During the tensile tests, before the gripper tightens a specimen, an invalid elongation about 1.1% in strain appears on the tensile curve. The length of a tensile curve from the post-yield point to the pre-fracture point is measured as the δ of the tensile specimen. Grain structures of the 2024 Al ingots and sheets were observed under polarized light. The samples were first grinded with a 600# abrasive paper, and then electrolytically polished at -40℃ ~ -30℃ with 20V DC in a solution of HNO3 : CH3OH = 2 : 8, at last anodized with Baker’s reagent (1% HBF4 in H2O) at 5~10℃ with 24V DC. SEM observation, EDS and EBSD analyses were conducted using a FEI Quanta-200 SEM instrument equipped with TSL OIMTM software. Samples for EBSD analysis were prepared according to the preparation procedure of metallographic samples without being anodized as last.
5
Results The grain structures of the CG and FG 2024 Al alloy ingots homogenized for 24h at 490℃ are shown in Fig. 2a and b, respectively, which are similar to the grain structures of the CG and FG ingots before homogenization, respectively. The large difference in grain size between the CG and FG ingots indicates the significant effectiveness of the combined grain refining technology of adding grain refiner and pouring at a low superheat. After hot rolling to the thickness reduction of 90%, the grains in both the two ingots are rolled into a long-pie-like morphology, as shown in Fig. 2c and d which show the grain structures of the hot rolled CG and FG sheets, respectively. After solution treatment for 1h at 500℃, significant recrystallization occurs in both the two sheets, as shown in Fig. 2e and f which show the grain structures of the CG and FG sheets after solution treat, respectively. Recrystallization results in a pan-cake-shaped grain morphology in the two sheets with little difference in grain size between the two sheets. This failure of refining the grain structure of the 2024 Al alloy sheet by refining the ingot grain structure puts forward the necessity of assessing the effect of refining ingot grain structure on other structural characteristics of the 2024 Al alloy sheet.
(c)
(d)
6
(f)
(e)
Fig. 2. Grain structures of (a, b) as-homogenized for 24h at 490℃ (c, d) hot rolled at 450℃ (e, f) solution-treated for 1h at 500℃ samples of (a, c, e) CG and (b, d, f) FG 2024 Al alloys.
Fig. 3a and b show the SEM images of the homogenized CG and FG 2024 Al alloy ingots, respectively, in which the white phases are the second phases remaining in the two ingots after homogenization. No significant difference in the remaining phase content between the two ingots can be found. In addition, EDS analysis reveals that the phase types in the two homogenized ingots are the same, indicating that the ingot grain refinement has no detectable effect on the content and types of remaining phases in the homogenized 2024 Al alloy ingot. Typical common remaining phases in the two ingots after homogenization are the complex AlCuFeMn phases and Al2Cu+Al2CuMg eutectic phases, as shown in Fig. 3c and d, respectively. Nonetheless, because the grains of the CG ingot are so large that some remaining phases in the ingot distributed between parallel dendrite arms beside a dendrite stem form a long phase lath, as shown at location A in Fig. 3a, and some remaining phases within a dendrite grain form parallel phase laths, as shown at location B in Fig. 3a, whereas in the FG ingot, all the remaining phases locate dispersedly on grain boundaries, as shown in Fig. 3b.
A B 7
B
A Fig. 3. SEM images of homogenized (a) CG and (b) FG 2024 Al ingots; (b, c) the common remaining phases in the two homogenized ingots; SEM images of solution treated (e) CG and (f) FG 2024 Al alloy sheets.
Fig. 3e and f show the SEM images of the CG and FG 2024 Al alloy sheets solution treated for 1h at 500℃, respectively. It can be seen that all the remaining phases in both the two sheets are
8
rolled into the particle chains lying parallel to their rolling planes. The first and most obvious difference is the more frequent occurrence of the long particle chains and parallel head-tail particle chains in the CG 2024 Al alloy sheet, as shown at locations A and B in Fig. 3e, respectively, which may be the preserved microstructural characteristics of the CG ingot as that at locations A and B in Fig. 3a, respectively. The second difference is the lower particle content in the FG sheet, even though the two homogenized ingots have approximately the same remaining phase contents, as indicated in Fig. 3a and b. This difference results from the more effective particle breakage occurring in the FG sheet during hot rolling, as will be analyzed later. Fig. 4a and b show the EBSD inverse pole figure (IPF) maps of the solution treated CG and FG 2024 Al alloy sheets, respectively, of which the EBSD image quality (IQ) maps are shown in Fig. 4c and d, respectively, in which the green lines represent the low-angle subgrain boundaries and the blue lines represent the high-angle grain boundaries formed during recrystallization. It is widely accepted that the subgrain boundaries formed during recovery of metals always have a misorientation angle lower than 15°and grain boundaries formed during recrystallization have a misorientation angle higher than 15°[27]. So by marking the subgrain boundaries and grain boundaries in the two sheets, the recrystallization extents in the two sheets can be evaluated. The presence of a large number of subgrains in Fig. 4c indicates that only a partial recrystallization occurs in the CG sheet after solution treatment, whereas the FG 2024 Al alloy sheet recrystallizes completely after solution treatment, as demonstrated in Fig. 4d. The IQ maps of the CG and FG 2024 Al alloy sheets without subgrain boundaries and grain boundaries being marked are shown in Fig, 4e and f, respectively. It can be found that in the CG sheet, those circled regions containing large particles have a completely recrystallized structure, whereas those regions with a recovered substructure have no large particles.
(a)
ND
(b)
RD
9
500μm
500μm
500μm
500μm
500μm Fig. 4. (a, b) EBSD IPF maps and (c, d) corresponding EBSD IQ maps, in which the green and blue lines represent the subgrain boundaries and grain boundaries, respectively, of (a, c) CG and (b, d) FG 2024 Al alloy sheets after solution treatment; (e) and (f) are the same IP maps as (c) and (d), respectively, without boundaries being marked.
10
All the above microstructure observation results indicate that even though the solution treated CG and FG 2024 Al alloy sheets have similar metallographic structures, their substructures and the contents and distributions of the remaining particles in them are different, which may then result in different tensile properties for the naturally aged CG and FG 2024 Al alloy sheets. The tensile properties of the two naturally aged sheets in different directions are shown in Fig. 5. First, neither of two sheets shows in-plane anisotropy, which should be attributed to their recrystallized grain structures, even though only a partial recrystallization has occurred in the CG sheet. Many researches [22, 28-32] concerning the mechanical properties of Al profiles have revealed that an unrecrystallized Al profile often shows an in-plane anisotropy of tensile properties with higher strength and lower ductility in the RD direction than that in the TD (90°) and 45° directions, whereas for the recrystallized Al profiles this in-plane anisotropy vanishes partially or even completely.
σ0.2 σb δ
CG sheet
δ\%
σ0.2, σb \ MPa
FG sheet
0°
45°
90°
ND
Fig. 5. Tensile properties of the naturally aged CG and FG 2024 Al alloy sheets in different directions.
Secondly, the tensile properties of the 0°, 45° and 90° samples of the CG sheet, σ0.2 ≈ 315MPa, σb ≈ 440MPa, δ ≈ 21%, are higher in strength and lower in ductility than that of the FG sheet, σ0.2 ≈ 280MPa, σb ≈ 420MPa, δ ≈ 27%. It is well known that the strength of the naturally aged 2024 Al alloy mainly originates from the age hardening [33]. Obviously in this study the subgrain strengthening mechanism also plays a part in the naturally aged CG sheet as a result of its non-completely recrystallized substructure. The subgrain strengthening mechanism can not only 11
increase an alloy’s strength and reduce its ductility accordingly, but reduce the difference between σ0.2 and σb of the alloy [30]. The difference is approximately 125MPa for the CG 2024 Al alloy sheet, and 140MPa for the FG 2024 Al alloy sheet, indicating that the higher strengths of the 0°, 45° and 90° samples of the CG sheet are caused by the finer substructure of the sheet. It should be noted, however, that even though there are more remaining particles in the CG 2024 Al alloy sheet, as shown in Fig. 3e, the strengths of the 0°, 45° and 90° samples of the CG sheet are not reduced compared with the strengths of the FG sheet in the three directions. This discrepancy has also been noted by other researchers [1, 34, 35]. These authors found that reducing the impurity particle contents in Al alloys by increasing their purities or employing different manufacture technologies cannot increase their strengths in the directions parallel to the preferential distribution planes of the impurity particles in these alloys, such as the rolling plane of a hot rolled Al alloy, but can increase their ductilities in these directions. In view of this, it is concluded here that the higher ductilities of the 0°, 45° and 90° samples of the FG 2024 Al alloy sheet are brought about together by their completely recrystallized substructures and lower impurity contents. At last, the most important result illustrated in Fig. 5 is that the FG 2024 Al alloy sheet has much higher strength and ductility in the ND direction, σ0.2 ≈ 280 MPa, σb ≈ 370 MPa, δ ≈ 7%, than that of the CG sheet, σ0.2 ≈ 260 MPa, σb ≈ 275 MPa, δ ≈ 3%. The differences in tensile properties between the two naturally aged 2024 Al alloy sheets can be further demonstrated from their fracture morphologies. The low magnification SEM images of the fracture surfaces of the CG and FG tensile specimens in the RD direction are illustrated in Fig. 6a and b, respectively, of which the higher magnification images are shown in Fig. 6c and d, respectively. It can be observed that except those regions containing large particles with a cleavage fracture morphology, the remaining fracture surfaces of the two sheets have a ductile fracture morphology with many dimples of 1 μm in size and containing small particles at their bottom. These small particles should be the T phase (Al20Cu2Mn3) in 2024 Al alloy [36]. It is evident that the two 2024 Al alloy sheets have the same facture mechanism, i.e., the ductile fracture mechanism, in the RD direction. It was also found in the experiments that the facture morphologies of the 45° and 90° samples of the two sheets are similar to what are shown in Fig. 6a and b.
12
13
Fig. 6. SEM micrographs of the fracture surfaces of (a, c, e) CG and (b, d, f) FG 2024 Al alloy sheets in the (a, b, c, d) RD and (e, f) ND directions.
The fracture morphologies of the CG and FG sheets in the ND direction are shown in Fig. 6e and f, respectively. It is evident that both the two ND tensile specimens fracture along the remaining particle chains in the two sheets, resulting in the abundance of large particles on their fracture surfaces. It has been observed [37, 38] that the large brittle impurity particles in Al alloys fracture at the early tensile stage, inducing crack sources in a specimen. In this study, as the remaining phases in the homogenized CG and FG ingots are hot rolled into the particle chains lying perpendicular to the ND directions of the two sheets, microcracks along these particle chains can be easily formed during the tensile tests in the ND direction, reducing the actual load-bearing areas in the two ND specimens. A longer particle chain will induce a longer microcrack which can unstably propagate more easily, so the more and longer particle chains present in the CG 2024 Al alloy sheet will dramatically reduce the strength and ductility of the sheet in the ND direction. In addition, it can also be observed from Fig. 6f that more and finer impurity particles are present on the fracture surface of the FG sheet, causing more and finer dimples accordingly than in the CG sheet. This fracture morphological difference between the two sheets evidences the conclusion that more effective particle breakage is caused in the FG sheet during the hot rolling, which is drawn above from the comparison of Fig. 3e and c. It has been reported [2] that the crack-originating capability of an impurity particle in Al alloys decreases dramatically as long as its size is reduced below 6μm, which can be interpreted qualitatively as that a microcrack formed with the fracture of a finer particle has a more weak stress concentration at its tip and so it will be more difficult to propagate. So another important reason for the higher ductility of the FG sheet in the ND direction is the finer remaining particles in it.
(a)
CG sheet FG sheet
14
Elongation \ %
(a)
CG sheet FG sheet
Strain \ %
15
Fig. 7. (a) Tensile stress-strain curves of the CG and FG 2024 Al alloy sheets in the ND direction; side profiles of the corresponding fracture surfaces of (b) CG and (c) FG 2024 Al alloy sheets.
The stress-strain curves of the CG and FG sheets in the ND direction are shown in Fig. 7a, and the side profiles of corresponding fracture surfaces are shown in Fig. 7b and c, respectively. From Fig. 7a it can be seen that the ND specimen of the CG sheet fractures not long after the yield point, without obvious strain hardening before fracture. This should be mainly attributed to the more and longer particle chains in the specimen which can induce more and longer microcracks of which the tip stress concentration is so high that, before significant strain hardening occurs in the metal matrix, the tensile stress has reached the level for the microcracks to unstably propagate. So in the ND specimen of the CG sheet the microcracks propagate much earlier during the tensile tests, reducing the stress abruptly as a result of their cleavage fracture mechanism as reflected in Fig. 6e. In the ND specimen of the FG sheet, the particle chains are less and shorter that not only the microcracks formed along the chains are more difficult to propagate as a result of their lower tip stress concentration, but also these microcracks are more difficult to be connected during propagation to form longer cracks as a result of their larger spacings. So a higher degree of plastic deformation can occur in the ND specimen of the FG sheet to cause a higher level of strain hardening in the metal matrix so as to promote the microcracks to unstably propagate, causing the more ductile fracture morphology of the specimen, as shown in Fig. 6f. It is also those longer particle chains in the CG sheet that lead to the straighter side profile of the fracture surface of the CG specimen in the ND direction, as shown in Fig. 7b, whereas a more curved side profile is formed for the fractured FG specimen, as shown in Fig. 7c, because more frequent crack-direction 16
changes are needed to connect the shorter microcracks formed along the shorter particle chains in the specimen.
Discussion All the above results indicate that the differences in the substructure and the content and distribution of the remaining phases between the two naturally aged sheets lead to slightly lower strengths and higher ductilities of the 0°, 45° and 90° samples and markedly higher strength and ductility of the ND sample for the FG sheet. The microstructural differences between the two 2024 Al alloys sheets not only have something to do with their ingot microstructures which are preserved to some extent in the two sheets after hot rolling, but also are influenced by the different degrees of deformation homogeneities of the two ingots during hot rolling.
(c)
(d)
RD
RD TD
TD
17
Fig. 8. Grain structures of (a) CG and (b) FG 2024 Al alloy sheets hot rolled by a thickness reduction of 60%; IPF maps of the rolling planes of (c) CG and (d) FG 2024 Al alloy sheets hot rolled by a thickness reduction of 90%; (e) and (f) are the corresponding SEM images of (c) and (d), respectively. (g) edge cracking of the two hot rolled sheets at the thickness reduction of 90%.
Microstructures of the CG and FG sheets hot rolled at the thickness reduction of 60% are shown in Fig. 8a and b, respectively. Deformation heterogeneity is evident in the CG sheet, with some severely deformed grains with deformation bands and some slightly deformed grains without obvious metallographic deformation structure, whereas all the grains in the FG sheet are flatly rolled to a similar degree that no obvious deformation heterogeneity can be found in the sheet. With deformation heterogeneity in the CG sheet, the breakage of a grain into different orientation regions during hot rolling is more evident in the sheet, resulting in reduced grain size difference between the hot rolled CG and FG sheets as shown in Fig. 2c and d, respectively, compared to the grain size difference between the CG and FG ingots shown in Fig. 2a and b, respectively. Fig. 8c and d show the IPF maps on the rolling planes of CG and FG sheets at the thickness reduction of 90%, respectively. Fig. 8e and f show the corresponding SEM images in
18
which the circled holes are where the large remaining particles locate which are preferentially polished during the sample preparation. Other smaller holes on the sample surfaces are the defects caused after the precipitated particles (Al2CuMg) formed during the hot rolling are preferentially polished. It can be found in Fig. 8c and d that significant recrystallization occurs in the two sheets during the hot rolling around the large particles. Particles in Al alloys above a critical size are effective to stimulate recrystallization during the hot working [39]. The recrystallization grains in the CG sheet are larger than that in the FG sheet, as can be seen in Fig. 8c and d. Recrystallization during hot rolling can cause recrystallization softening, inducing strain concentration in these recrystallized regions. As a result, the strain concentration should be higher in the CG sheet as a result of the larger recrystallization grains in the sheet, and so the unrecrystallized regions in the CG sheet will experience lower deformation than that in the FG sheet. This conclusion is supported by the result that higher orientation variation is present in the unrecrystallized regions in the FG sheet, as shown in Fig. 8c and d, suggesting higher degree of deformation heterogeneity in the CG sheet. The edge cracking of the two sheets at the thickness reduction of 90% are shown in Fig. e. It is evident that the cracks on the CG sheet edge are longer, deeper and are more unevenly distributed. The higher susceptibility of the CG sheet to the edge cracking has a relationship with the higher strain concentration in the recrystallized regions of the sheet, which was also demonstrated by others [40]. The higher degree of deformation heterogeneity also leads to a higher degree of stored energy heterogeneity in the CG sheet after hot rolling, which then leads to the non-complete recrystallization substructure in the sheet after solution treatment, as shown in Fig. 4e. This suggests that the recrystallization in the two sheets is affected not only by the large particles but also by their deformation homogeneities during hot rolling. According to some investigations [22, 23], increasing the hot working temperature of Al profiles will increase their recrystallization temperature accordingly. This is because a higher hot working temperature will results in a higher degree of dynamic recovery during the hot working, reducing the stored deformation energy in Al profiles and so increasing their recrystallization temperature. So it is suggested here that by controlling the hot working temperature of an FG Al alloy ingot, the substructure of its profile after the ultimate heat treatment can be chosen purposely to be the completely recrystallized or recovered one, which is useful for the application requirement that the deformation structure of an 19
Al profile needs to be preserved after the ultimate heat treatment so as to preserve its higher strength in the rolling or extrusion direction. In this case, adapting an Al profile manufactured from the FG ingots with recovered substructure can not only prevent the recrystallization softening, but endow the profile with higher ductility as a result of the reduced size and improved distribution of the remaining large particles in the Al profile. For an Al profile manufactured from a CG Al ingot it is more difficult to control its substructure by controlling the hot working temperature as a result of its heterogeneous deformation substructure. For the on-line quenching Al profiles, a finer ingot grain structure will bring about a finer microstructure for the Al profiles, as shown in Fig. 2d, which is evidently beneficial for improving their mechanical properties. It can be observed from Fig. 3a and b that some of the remaining phases in the CG ingot distribute between dendrite arms, i.e., within the grains, whereas all the remaining phases in the FG ingot locate at grain boundaries. It is well known that during the deformation of metals the regions near grain boundaries experience more severe deformation than that within grains [41]. This will result in a more effective breakage of the remaining phases in the FG ingot. In addition, because some grains in the CG ingot experience a slightly weaker deformation, as shown in Fig. 8a and c, the remaining particles in these grains will be slightly broken. Both the two above conditions will result in the less effective particle breakage in the CG 2024 Al alloy sheet, as illustrated in Figs. 3e and 6e. The improved strength and ductility of the FG 2024 Al alloy sheet in the ND direction will endow the sheet with higher service reliability. It is shown in Fig. 9a the schematic diagram of the microstructure at the base location of an I type extruded Al profile [30]. First, a stress concentration is more likely to occur at this corner location. Second, the fibrous grain structure at P location leads the profile to be vulnerable to the splitting along grain boundaries containing impurity particles under a tensile stress either in the vertical or in the horizontal direction. In this case, it is particularly beneficial to use a FG Al alloy ingot to manufacture the Al profile so as to increase its strengths and ductilities in the vertical and horizontal directions.
20
(a) p
RD and Tensile axis
Fig. 9. (a) Schematic diagram of the fibrous grain structure at the base location of an I type extruded Al profile [30]; (b) side profile of the fracture surface of the CG 2024 Al alloy specimen in the RD direction. The shear failure surface of the specimen is disconnected by the splitting (indicated by red lines) along the remaining particle chains.
Fig. 9b shows the side profile of the fracture surface of the CG 2024 Al alloy specimen in the RD direction. It can be seen that even though its shear failure surfaces are inclined at 45° to the tensile direction, those microcracks in the RD direction (indicated by red lines), which should be caused by the splitting along the particle chains in the specimen, disconnect the main crack onto different shear planes. It was reported that [42], even without reducing the strength of an Al profile, this influence of impurity particles on the crack propagation can reduce the toughness of the Al profile by rendering a Zig-Zag or twisted-type of crack propagation pattern. In addition, reducing the size and number of the impurity particles in Al profiles is beneficial for improving their fatigue properties [18, 43]. In view of this, FG Al alloy ingots are suggested to be used so as to improve the mechanical properties of Al profiles, because in current industrial practices, it is common to adopt Al alloy ingots with dendrite grains of several hundreds of μm in diameter to produce Al profiles.
21
Conclusions By employing a combined casting technology of adding 0.1 wt.% Al-5Ti-1B grain refiner and pouring at a low superheat (650℃), solidification structure of the 2024 Al alloy ingot is significantly refined. However, this grain refinement has little effect on the types and content of the remaining phases in the ingot after homogenization, but leads to a more even distribution of the remaining phases. The grain refinement also leads to a more homogenous deformation and lower degree of recrystallization in the FG sheet during hot rolling as a result of the finer recrystallization grains in the sheet, and hence lightens the edge-cracking of the sheet. Even though the grain sizes of the two solution treated 2024 Al alloy sheets are the same, a non-complete and complete recrystallization occur in the CG and FG sheets, respectively. With more dispersedly distributed remaining phases in the homogenized FG ingot and more effective breakage of the remaining particles during the hot rolling in the FG sheet. the remaining particles are smaller and the remaining particle chains are shorter and less in the FG sheet. These above microstructural differences between the two sheets result in slightly lower strengths and higher ductilities of the 0°, 45° and 90° samples and markedly higher strength and ductility of the ND sample for the FG sheet, which suggests that the benefits of ingot grain refinement should not be narrowly regarded as being able to improve the casting and hot working capabilities of an ingot.
References [1] J.S. Santner, A study of fracture in high purity 7075 aluminum alloys, Metall. Trans. A 9 (1978) 769-779. http://dx.doi.org/10.1007/BF02649786 [2] C.Y. Kung, M.E. Fine, Fatigue crack initiation and microcrack growth in 2024-T4 and 2124-T4 aluminum alloys, Metall. Trans. A 10 (1979) 603-610. http://dx.doi.org/doi:10.1007/BF02658324 [3] N.E. Prasad, S.V. Kamat, K.S. Prasad, G. Malakondaiah, V.V. Kutumbarao, In-plane anisotropy in the fracture toughness of an Al-Li 8090 alloy plate, Eng. Fract. Mech. 46 (1993) 209-223. http://dx.doi.org/10.1016/0013-7944(93)90282-W [4] The effect of grain size on low-cycle fatigue behavior of Al-2024 polycrystalline alloy, Metall. Mater. Trans. A 35 (2004) 2725-2728. http://dx.doi.org/10.1007/s11661-004-0218-1 [5] T. Warner, Recently-developed aluminium solutions for aerospace applications, Mater. Sci. Forum. 519 (2006) 1271-1278. http://dx.doi.org/10.4028/www.scientific.net/MSF.519-521.127 [6] J. Staley, Influence of microstructure on fatigue and fracture of aluminum alloys, Aluminum 55 (1979) 277-281. 22
[7] N.E. Paton, C.H. Hamilton, J.A. Wert, M.W. Mahoney, Characterization of fine-grained superplastic aluminum alloys, JOM 34 (1982) 21-27. http://dx.doi.org/doi:10.1007/BF03338068 [8] J. Waldman, H. Sulinski, H. Markus, The effect of ingot processing treatments on the grain size and properties of Al alloy 7075, Metall. Trans. 5 (1974) 573-584. http://dx.doi.org/10.1007/bf02644652 [9] B.S. Murty, S.A. Kori, M. Chakraborty, Grain refinement of aluminium and its alloys by heterogeneous
nucleation
and
alloying,
Int.
Mater.
Rev.
47
(2002)
3-29.
http://dx.doi.org/10.1179/095066001225001049 [10] T. Campanella, C. Charbon, M. Rappaz, Grain refinement induced by electromagnetic stirring: A dendrite
fragmentation
criterion,
Metall.
Mater.
Trans.
A
35
(2004)
3201-3210.
http://dx.doi.org/10.1007/s11661-004-0064-1 [11] M.A. Easton, H. Kaufmann, W. Fragner, The effect of chemical grain refinement and low superheat pouring on the structure of NRC castings of aluminium alloy Al–7Si–0.4Mg, Mater. Sci. Eng. A 420 (2006) 135-143. http://dx.doi.org/10.1016/j.msea.2006.01.078 [12] E. Wang, T. Gao, J. Nie, X. Liu, Grain refinement limit and mechanical properties of 6063 alloy inoculated
by
Al–Ti–C
(B)
master
alloys,
J.
Alloy.
Compd.
594
(2014)
7-11.
http://dx.doi.org/10.1016/j.jallcom.2014.01.145 [13] A.B. Pattnaik, S. Das, B.B. Jha, N. Prasanth, Effect of Al–5Ti–1B grain refiner on the microstructure, mechanical properties and acoustic emission characteristics of Al5052 aluminium alloy, J. Mater. Res. Technol. 4 (2015) 171-179. http://dx.doi.org/10.1016/j.jmrt.2014.10.017 [14] S. Meenia, F. Khan Md, S. Babu, R.J. Immanuel, S.K. Panigrahi, G.D. Janaki Ram, Particle refinement and fine-grain formation leading to enhanced mechanical behaviour in a hypo-eutectic Al–Si alloy subjected to multi-pass friction stir processing, Mater. Charact. 113 (2016) 134-143. http://dx.doi.org/10.1016/j.matchar.2016.01.011 [15] L. Bolzoni, M. Nowak, N. Hari Babu, Grain refinement of Al–Si alloys by Nb–B inoculation. Part II: Application
to
commercial
alloys,
Mater.
Design.
66
(2015)
376-383.
http://dx.doi.org/10.1016/j.matdes.2014.08.067 [16] R. Takai, S. Kimura, R. Kashiuchi, H. Kotaki, M. Yoshida, Grain refinement effects on the strain rate sensitivity and grain boundary sliding in partially solidified Al–5 wt%Mg alloy, Mater. Sci. Eng. A 667 (2016) 417-425. http://dx.doi.org/10.1016/j.msea.2016.05.023 [17] P.N. Anyalebechi, Characterization of the Microstructure of Commercial-Size Ingots of Aluminum Alloy
3004,
EPD
Congress
2011,
John
Wiley
&
Sons,
Inc.,
pp.
678-700.
http://dx.doi.org/10.1002/9781118495285.ch75 [18] J.H. Mulherin, H. Rosenthal, Influence of nonequilibrium second-phase particles formed during solidification upon the mechanical behavior of an aluminum alloy, Metall. Trans. 2 (1971) 427-432. http://dx.doi.org/10.1007/bf02663329 [19] W. Zhang, G.S. Frankel, Transitions between pitting and intergranular corrosion in AA2024, Electrochim. Acta. 48 (2003) 1193-1210. http://dx.doi.org/10.1016/S0013-4686(02)00828-9 [20] W. Shuncheng, G. Chunlei, L. Xiaohui, Z. Kaihong, Q. Wenjun, Microstructure of Al-4.99Zr-1.1B Master Alloy and Its Grain Refinement Effect on AZ31 Magnesium Alloy, Rare. Metal. Mat. Eng. 43 (2014) 2567-2571. http://dx.doi.org/10.1016/S1875-5372(15)60002-9 [21] J. Mizera, J.H. Driver, E. Jezierska, K.J. Kurzydłowski, Studies of the relationship between the microstructure and anisotropy of the plastic properties of industrial aluminum-lithium alloys, Mater. Sci. Eng. A 212 (1996) 94-101. http://dx.doi.org/10.1016/0921-5093(96)10172-6 [22] K.V. Jata, S. Panchanadeeswaran, A.K. Vasudevan, Evolution of texture, micro structure and
23
mechanical property anisotropy in an Al-Li-Cu alloy, Mater. Sci. Eng. A 257 (1998) 37-46. http://dx.doi.org/10.1016/S0921-5093(98)00822-3 [23] K. TAKAHASHI, K. MINAKAWA, C. OUCHI, The effect of thermomechanical processing variables on anisotropy in mechanical properties of Al-Li alloys, J. Phys. Colloques. 48 (1987) C3-163-C163-169. http://dx.doi.org/10.1051/jphyscol:1987319 [24] A. Heinz, A. Haszler, C. Keidel, S. Moldenhauer, R. Benedictus, W.S. Miller, Recent development in aluminium alloys for aerospace applications, Mater. Sci. Eng. A 280 (2000) 102-107. http://dx.doi.org/10.1016/S0921-5093(99)00674-7 [25] G. Bussu, P.E. Irving, The role of residual stress and heat affected zone properties on fatigue crack propagation in friction stir welded 2024-T351 aluminium joints, Int. J. Fatigue. 25 (2003) 77-88. http://dx.doi.org/10.1016/S0142-1123(02)00038-5 [26] R. Zettler, S. Lomolino, J.F. dos Santos, T. Donath, F. Beckmann, T. Lippman, D. Lohwasser, Effect of Tool Geometry and Process Parameters on Material Flow in FSW of an AA 2024-T351 Alloy, Weld World 49 (2005) 41-46. http://dx.doi.org/10.1007/bf03266474 [27] S. Gourdet, F. Montheillet, An experimental study of the recrystallization mechanism during hot deformation
of
aluminium,
Mater.
Sci.
Eng.
A
283
(2000)
274-288.
http://dx.doi.org/10.1016/S0921-5093(00)00733-4 [28] R.E. Zinkham, Anisotropy and thickness effects in fracture of 7075-T6 and -T651 aluminum alloy, Eng. Fract. Mech. 1 (1968) 275-276. http://dx.doi.org/10.1016/0013-7944(68)90003-9 [29] M.H. SKILLINGBERG, R.F. ASHTON, Processing and performance of Al-Li-Cu-X extrusions, J. Phys. Colloques 48 (1987) C3-179-C173-186. http://dx.doi.org/10.1051/jphyscol:1987321 [30] S.J. Hales, R.A. Hafley, Texture and anisotropy in Al-Li alloy 2195 plate and near-net-shape extrusions, Mater. Sci. Eng. A 257 (1998) 153-164. http://dx.doi.org/10.1016/S0921-5093(98)00834-X [31] R.K. Singh, A.K. Singh, N.E. Prasad, Texture and mechanical property anisotropy in an Al–Mg–Si– Cu alloy, Mater. Sci. Eng. A 277 (2000) 114-122. http://dx.doi.org/10.1016/S0921-5093(99)00549-3 [32] W. Hasso, Industrial application of recrystallization control in aluminum products, Mater. Sci. Forum. 467 (2004) 349-356. http://dx.doi.org/10.4028/www.scientific.net/MSF.467-470.349 [33] R.K.W. Marceau, G. Sha, R. Ferragut, A. Dupasquier, S.P. Ringer, Solute clustering in Al–Cu–Mg alloys during the early stages of elevated temperature ageing, Acta. Mater. 58 (2010) 4923-4939. http://dx.doi.org/10.1016/j.actamat.2010.05.020 [34] P.J. Blau, Influence of Iron and Silicon Content on the Tensile Properties of 7X75 and ZR-Modified 7X75 Aluminum Plate, Air Force Materials Lab., Wright-Patterson AFB, Ohio (USA). [35] D.S. Thompson, Metallurgical factors affecting high strength aluminum alloy production, Metall. Trans. A 6 (1975) 671-683. http://dx.doi.org/10.1007/BF02672287 [36] Y.Q. Chen, D.Q. Yi, Y. Jiang, B. Wang, D.Z. Xu, S.C. Li, Twinning and orientation relationships of T-phase
precipitates
in
an
Al
matrix,
J.
Mater.
Sci.
48
(2013)
3225-3231.
http://dx.doi.org/10.1007/s10853-012-7102-y [37] G.T. Hahn, A.R. Rosenfield, Metallurgical factors affecting fracture toughness of aluminum alloys, Metall. Trans. A 6 (1975) 653-668. http://dx.doi.org/10.1007/bf02672285 [38] R.H. Van Stone, J.A. Psioda, Discussion of “metallurgical factors affecting fracture toughness of aluminum alloys”, Metall. Trans. A 6 (1975) 668-670. http://dx.doi.org/10.1007/bf02672286 [39] J.D. Robson, Microstructural evolution in aluminium alloy 7050 during processing, Mater. Sci. Eng. A 382 (2004) 112-121. http://dx.doi.org/10.1016/j.msea.2004.05.006 [40] J.J. Sundelin, S.T. Nurmi, T.K. Lepistö, Recrystallization behaviour of SnAgCu solder joints, Mater.
24
Sci. Eng. A 474 (2008) 201-207. http://dx.doi.org/10.1016/j.msea.2007.04.013 [41] W.Z. Abuzaid, M.D. Sangid, J.D. Carroll, H. Sehitoglu, J. Lambros, Slip transfer and plastic strain accumulation across grain boundaries in Hastelloy X, J Mech. Phys. Solids 60 (2012) 1201-1220. http://dx.doi.org/10.1016/j.jmps.2012.02.001 [42] H.X. Li, C.Q. Chen, Mechanism of anisotropy in fracture behaviour and fracture toughness of high strength
aluminium
alloy
plate,
Mater.
Sci.
Tech-lond.
6
(1990)
850-856.
http://dx.doi.org/10.1179/mst.1990.6.9.850 [43] J. Grosskreutz, G. Shaw, D. Benson, The effect of inclusion size and distribution on fatigue of 2024-T4 aluminum, Wright-Patterson Air Force Base, Dayton, Ohio, USA.
25
Figure captions Fig. 1. (a) Mini-tensile specimen dimension; (b) the gripper matching the mini-tensile specimen; (c) schematic of different specimen orientations; (d) stress-strain curve of a mini-tensile specimen, to which the necking occurs mainly within its gauge region. During the tensile tests, before the gripper tightens a specimen, an invalid elongation about 1.1% in strain appears on the tensile curve. The length of a tensile curve from the post-yield point to the pre-fracture point is measured as the δ of the tensile specimen. Fig. 2. Grain structures of (a, b) as-homogenized for 24h at 490℃ (c, d) hot rolled at 450℃ (e, f) solution-treated for 1h at 500℃ samples of (a, c, e) CG and (b, d, f) FG 2024 Al alloys. Fig. 3. SEM images of homogenized (a) CG and (b) FG 2024 Al ingots; (b, c) the common remaining phases in the two homogenized ingots; SEM images of solution treated (e) CG and (f) FG 2024 Al alloy sheets. Fig. 4. (a, b) EBSD IPF maps and (c, d) corresponding EBSD IQ maps, in which the green and blue lines represent the subgrain boundaries and grain boundaries, respectively, of (a, c) CG and (b, d) FG 2024 Al alloy sheets after solution treatment; (e) and (f) are the same IP maps as (c) and (d), respectively, without boundaries being marked. Fig. 5. Tensile properties of the naturally aged CG and FG 2024 Al alloy sheets in different directions. Fig. 6. SEM micrographs of the fracture surfaces of (a, c, e) CG and (b, d, f) FG 2024 Al alloy sheets in the (a, b, c, d) RD and (e, f) ND directions. Fig. 7. (a) Tensile stress-strain curves of the CG and FG 2024 Al alloy sheets in the ND direction; side profiles of the corresponding fracture surfaces of (b) CG and (c) FG 2024 Al alloy sheets. Fig. 8. Grain structures of (a) CG and (b) FG 2024 Al alloy sheets hot rolled by a thickness reduction of 60%; IPF maps of the rolling planes of (c) CG and (d) FG 2024 Al alloy sheets hot rolled by a thickness reduction of 90%; (e) and (f) are the corresponding SEM images of (c) and (d), respectively. (g) edge cracking of the two hot rolled sheets at the thickness reduction of 90%. Fig. 9. (a) Schematic diagram of the fibrous grain structure at the base location of an I type extruded Al profile [30]; (b) side profile of the fracture surface of the CG 2024 Al alloy specimen in the RD direction. The shear failure surface of the specimen is disconnected by the splitting (indicated by red lines) along the remaining particle chains.
26