- Email: [email protected]
Effect of welding thermal cycle on microstructural evolution of Al–Zn–Mg–Cu alloy
Author’s Accepted Manuscript Effect of welding thermal cycle on microstructural evolution of Al–Zn–Mg–Cu alloy Kang Zhang, J.Q. Chen, P.Z. Ma, X.H. Zhang
www.elsevier.com/locate/msea
PII: DOI: Reference:
S0921-5093(18)30096-0 https://doi.org/10.1016/j.msea.2018.01.067 MSA36021
To appear in: Materials Science & Engineering A Received date: 23 October 2017 Revised date: 17 January 2018 Accepted date: 18 January 2018 Cite this article as: Kang Zhang, J.Q. Chen, P.Z. Ma and X.H. Zhang, Effect of welding thermal cycle on microstructural evolution of Al–Zn–Mg–Cu alloy, Materials Science & Engineering A, https://doi.org/10.1016/j.msea.2018.01.067 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 welding thermal cycle on microstructural evolution of Al–Zn–Mg–Cu alloy Kang Zhanga, J. Q. Chena,*, P. Z. Maa, X. H. Zhanga a Key Laboratory of Advanced Technologies of Materials, Ministry of Education, School of Materials Science and Engineering, Southwest Jiaotong University, Chengdu 610031, China * [email protected] ABSTRACT Al–Zn–Mg–Cu alloys are used extensively in high-speed train applications; however, the occurrence of softening in the heat-affected zone after welding limits their development. In this work, the effect of the welding thermal cycle on the softening and aging behaviors of a 7xxx aluminum alloy was investigated by examining the state of the strengthening precipitates. The microstructure and solvus temperature range of the precipitates of 7N01P-T4 were characterized using TEM and DSC analysis, respectively. It was found that the softening behavior of the aluminum alloys was closely related to the volume fraction and size of the hardening precipitates, which were greatly affected by the peak temperature of the welding thermal cycle. In addition, η′ and η precipitates were observed to be primarily responsible for the increase in the mechanical and electrical properties during the room-temperature natural aging. A phenomenological connection was thus uncovered between the characteristic parameters of the thermal cycle and the precipitation behavior, providing insight for the design of the welding process for 7N01 alloys. Keywords: Al–Zn–Mg–Cu alloy, thermal cycle, softening behavior, TEM, precipitation 1. Introduction With the trend of the development of light-weight structures, high-strength aluminum alloys have been used extensively. Alloy 7N01, a heat-treatable Al–Mg–Zn alloy, was developed for train body applications in the high-speed train industry [1]. In these applications, the alloys must be joined using a welding process; therefore, the effect of welding on the material properties is an important consideration. Unfortunately, softening generally occurs in the heat-affected zone (HAZ) of heat-treatment-strengthened Al alloys during the welding process. In addition, the location of the softened zone is typically a potential fatigue failure location during in-service processes, which seriously affects the safety of the structure, reported by Grong [2]. Therefore, in the present research, the coupling of dissolution and precipitation in the HAZ and the peak temperature in the thermal cycle was investigated, as it is important for clarifying the softening problem in the 7N01 welding process. Unlike heat treatment, a welding thermal cycle is a rapid heating and cooling process. The dissolution and precipitation of strengthening phases in the HAZ occur during the welding cycle and subsequent natural aging. Bjørneklett et al. [3] used process modeling techniques to develop a model that provides a sound basis for predicting the HAZ strength loss during welding of Al–Zn–Mg alloys as well as the resulting recovery of strength after prolonged natural aging. Because the HAZ of the weld was too narrow to prepare test specimens, Sato et al. [4] performed thermal simulation to evaluate specimens with a wide uniform temperature zone, and subsequent transmission electron microscopy (TEM) observation revealed that the precipitate distribution corresponded to the hardness profile in the HAZ of Al alloys. Hansen et al. [5] evaluated the effect of time–temperature ranges of the precipitate transformations of 7000 series alloys during the aging process. In addition, the dependence on the quenching temperature was analyzed using differential scanning calorimetry (DSC) and TEM. The mechanical properties of precipitation-strengthened Al alloys were observed to be closely related to the volume fraction and size of the hardening precipitates. Zhang et al. [6] monitored the evolution of the 1
precipitate size with progressing deformation using bright-field TEM images. The statistical results revealed an association between the microstructural evolution and mechanical behavior of an aged 6xxx alloy with finely dispersed η′ precipitates. In addition, small-angle X-ray scattering (SAXS) was recently used for an in situ quantitative study at elevated temperature of the mean size and fraction of precipitates in a 7xxx alloy [7,8,9], Hutchinson et al. [10] used SAXS for an in situ, elevated temperature study of the effect of monotonic strain and strain rate on the precipitate distribution in a 7xxx alloy，their results show that SAXS is mostly sensitive to the measurement of the precipitation state. However, because of the limitation of the heating mode and synchrotron radiation, studies of the evolution of precipitates during the welding of aluminum alloys using SAXS have seldom been reported. These considerations point to the need for a statistical methodology to obtain an in-depth understanding of precipitates, namely a quantitative measure of the mean size and fraction of precipitates for investigation of the softening problem for Al–Zn–Mg–Cu alloys. DSC and TEM are conventional methods that meet this requirement. Characterization of the solvus temperature range and mean radii of precipitates by DSC and TEM, respectively, can be used to examine the relationship between strengthening precipitates and thermal cycles. However, studying thermal cycle–precipitation coupling using this method slightly reduces the precision of the observations, partly because TEM itself can only be used to examine the micro-regional regularity of small precipitates and because the dislocations around precipitates make TEM observations difficult. The purpose of this study was not only to reveal the relationship between hardened precipitates and the thermal cycle but also to complement studies that have previously examined whether the physical performance of 7N01 alloy could be improved during the natural aging period. In this case, the use of TEM and DSC was acceptable, and the HAZ samples were processed using the thermal simulation technique. Hence, the specific purpose of this work was to investigate (i) the effect of thermal cycling on precipitates of a 7xxx alloy and (ii) the evolution of physical properties of materials (after being processed by thermal cycling) during a natural aging period. The generally poor mechanical performance of the 7xxx Al alloys was evaluated directly after the thermal cycle, and precipitate strengthening was considered after subsequent natural aging. This study of the HAZ softening problem will provide insight for the design of a more reasonable welding process. 2. Materials and experimental preparation 2.1 Temperature measurement during welding process In this study, the peak temperature of the thermal cycle in the HAZ of 7N01P-T4 aluminum alloy was first characterized. The chemical composition of 7N01P-T4 is provided in Table 1. Table 1 Chemical composition of aluminum alloy 7N01P-T4 (wt. %) material
Si
Fe
Cu
Mn
Mg
Cr
Ni
Zn
Ti
V
Zr
Al
7N01P-T4
＜0.001
0.087
0.032
0.377
1.049
0.089
＜0.001
4.23
0.029
0.014
0.083
Bal.
The HAZ of the 7N01 alloy was subjected to thermal cycling using one-pass hybrid laser–metal inert gas (MIG) welding with the welding parameters listed in Table 2. During the welding process, a pair of K-type thermocouples was used to monitor the temperature evolution of the thermal cycle. The thermocouples were arranged as shown in Fig. 1 at different distances (0, 3, 5, 7, 10, and 15 mm) from fusion line in the transverse direction. Table 2 Parameters of hybrid laser–MIG welding Laser power Wire speed Speed welding Welding voltage Welding current (Kw) (m/min) (m/min) (V) (A) 2
2
9
0.72
22.4
240
Fig. 1. Schematic diagram of welded plate used for measurement of thermal cycle showing the layout of the thermocouples 2.2 Mechanical property measurement
Fig. 2. Schematic diagram of specimen for thermal cycle simulation (specimens were cut with the longitudinally along the rolling direction) (a) tensile test specimen, (b) hardness and electrical conductivity specimen Firstly, the simulated thermal cycles (acquired by thermocouples) were loaded on the specimens by Gleeble machine. After being processed using the different thermal cycles, the Vickers hardness (3 kg load held for 10 s), electrical conductivity (%IACS)( as a supplement to the hardness to characterize the strengthening phase changes after thermal cycle) and tensile test (strain rate of 3mm/min) were measured once a week to monitor the softening behavior of the alloy with natural aging. The test specimen for the thermal–mechanical simulator is shown as Fig. 2. The hardness, electrical conductivity and tensile test were performed with a HVS-30 tester, a 7501 eddy current conductivity meter and a Gleeble machine, respectively. The mean value of ten measurement data was taken as the hardness and the electrical conductivity of the corresponding condition. 2.3 DSC and TEM observations DSC measurements were performed using a TA Instruments Discovery DSC calorimeter at heating 3
rates of 5°C/min, 25°C/min, and 50°C/min, respectively. High-purity aluminum of the same weight was used as a reference. During the test, the chamber was evacuated and then backfilled with pure argon gas. The microstructural changes occurring in the softening zone with the thermal cycling were examined using TEM. Thin foils specimens 30μm in thickness were obtain from the plates (Fig.2. (b)) for TEM observation. Electron-transparent sections of thin foils were prepared using a twin-jet electropolisher with a mixture of 70% methanol and 30% nitric acid (−25°C with liquid nitrogen). An FEI Titan G2 60-300 microscope operating at 200 kV was used for the TEM observation. The crystal models were built by Crystal Maker software. The measurements of the TEM images were done with the Image J software and Digital Micrograph software. 3. Results and discussion 3.1 Softening behavior after welding thermal cycles According to previous research [11], the HAZ of 7xxx alloys consists of two parts: a solid-solution zone and an overaging zone. The width of the HAZ in one-pass welding of 7N01 is approximately 3–10 mm, which corresponds to peak temperatures of approximately 150°C–550°C, as shown in Fig. 3, which were used to the primary peak temperatures of the simulated thermal cycle. First, thermal cycles at different peak temperatures of 150°C, 200°C,300°C, 400°C, and 500°C were simulated using a Gleeble-3500 thermal–mechanical simulator. Corresponding to the temperarure measurement during welding process, the specimen was heated to the peak temperature at a rate of 150°C/s and held at the peak temperature for 2 s, then air cooled to room temperature, as illustrated in Fig. 4. Tensile tests at room temperature were performed immediately after the samples were subjected to welding thermal cycles. Figure 5 reveals the severe decline in strength of the aluminum alloys after being subjected to thermal cycles with peak temperatures between 300°C and 500°C. (initial.T4 in Fig.5 stands for the initial state of the material without being subjected to thermal cycle ). This finding indicates that a large number of precipitates dissolved in this temperature range. Thus, peak temperatures of 300°C, 400°C, and 500°C were used to study the mechanical properties of the samples during the late natural aging period.
Fig. 3. Peak temperature distribution of 7N01 alloy measured by thermocouple during welding experiment
4
Fig. 4. The thermal cycles applied to 7N01 alloy
Fig. 5. Stress–strain curves of 7N01 alloy after being subjected to welding thermal cycles with different peak temperatures by the thermal–mechanical simulator 3.2 DSC analysis of the precipitate transition of 7N01P-T4
5
Fig. 6. DSC curves of 7N01P-T4 at different heating rates To better understand the specific relationship between the thermal cycle and precipitates, DSC tests were conducted. The effect of the heating rate on the initial state of 7N01P-T4 alloy is shown in Fig. 6. The number of peaks in the DSC curve is related to the nature of the material. The curves consist of wide and sharp peaks (endothermic and exothermic), with the location of the peaks changing depending on the heating rate. The differences in the peaks in Fig. 6 are considered to be mainly determined by the equilibrium phase transition suppressed , which causes the hysteresis effect of reaction. Fig. 6 further reveals that the heat effect of precipitate dissolution, formation, and growth depends on the heating rate. According to previous research reported by Deschamps et al. [12], the first peak of the DSC curve corresponds to GP zone formation, which generally occurs below 150°C in Al–Zn–Mg–Cu alloys; however, no significant endothermic or exothermic peak due to GP zone formation or dissolution was observed in the present alloys. This finding may indicate that the cluster and GP zone formation was completed in the present state of the alloy. As observed in Fig. 6, Peak I (the first endothermic peak) is a combination of two main reactions, mostly the dissolution of η′ and fine η. Peak II (the first exothermic peak) represents the formation and coarsening of η and corresponds to the determination of the activation energy of η. The amounts of η′ and η in the alloy depend on the Mg and Zn contents in 7N01, which, as a Zn-rich Al–Zn–Mg alloy, has a greater tendency to form precipitates described in Buha et al. [13]. Peak III (the second endothermic peak) corresponds to the dissolution of η. Because of the hysteresis effect caused by the high heating rate, the transformation temperature of the endothermic peak is higher, thereby delaying the corresponding dissolution effect during the welding thermal cycles in the HAZ. Combined with the observations in Fig. 5, a serious decline in strength occurred after the thermal cycles with peak temperatures between 300°C and 550°C, as determined by the dissolution of η′ and the formation/coarsening of η. Hence, Fig. 7 shows the dissolution and reprecipitation sequence of precipitates during the welding thermal cycles in the HAZ. In the initial T4 state of 7N01, precipitates are homogeneously distributed with sufficient growth. During the welding thermal cycle, dissolution of GP zones, η′, and fine η occurred during the heating period, and the dissolution of solute atoms back into the Al matrix occurred during the cooling period.
Fig. 7. Schematic diagram of dissolution during welding thermal cycle 6
3.3 Mechanical property of 7N01P-T4 alloy during the following natural aging
Fig. 8. Effect of natural aging and peak temperature of the thermal cycle on the (a) vicker’s hardness; (b) 0.2% proof strength of 7N01 aluminum alloy; (c) electrical conductivity. In the present study, the effect of the peak temperature of the thermal cycle on the mechanical and electrical properties was investigated during natural aging. To evaluate the mechanical properties of 7N01 aluminum alloy during natural aging, hardness and tensile tests were performed. With heat treatment condition change, it exhibits a correlation between the conductivity and mechanical properties, which could reflect the state of strengthening phase. Therefore, the electrical conductivity was measured to evaluate the transformation of precipitation and stress corrosion cracking (SCC) resistance of the alloy during natural aging[14,15,16]. As observed in Fig. 8, the hardness, strength and electrical conductivity of the 7N01 alloy after the welding thermal cycle are plotted as a function of the natural aging time. Increasing the natural aging time clearly led to a decrease of the electrical conductivity and increase of the hardness and strength. Moreover, it was concluded that a higher peak temperature leads to a lower electrical conductivity and hardness when the curve tends to be stable. A similar trend was observed for the strength. One week after the welding thermal cycle, significant softening to the lowest hardness value was observed, with a value significantly lower than that of the initial T4 state (124HV). Similarly, the strength was lower than that of the the initial T4 state (313 MPa). In contrast, the conductivity increased significantly (30%IACS of the T4 state), which was caused by the dissolution of precipitates and the re-dissolution of a large number of solute atoms into the Al matrix. The Matthiessen–Fleming rule is typically used to describe the electrical resistivity (the reciprocal of electrical conductivity) of materials. However, the electrical resistivity of multiphase alloy is relatively complex because of the effect of multiple precipitates, and no consensus regarding its behavior has been reached. A simple and modified Matthiessen rule can be used to describe the effect of precipitates on 7
electrical resistivity [17]: 𝑗
𝑖 𝑖 𝜌 = 𝜌𝑚 + ∑𝑖 𝑐𝑠𝑜𝑙𝑢𝑡 𝜌𝑠𝑜𝑙𝑢𝑡 + 𝑐𝑣𝑎𝑐 𝜌𝑣𝑎𝑐 + 𝑐𝑑𝑖𝑠 𝜌𝑑𝑖𝑠 + 𝑐𝐺𝐵 𝜌𝐺𝐵 + ∑𝑗 𝜌𝑝𝑟𝑒𝑐𝑖𝑝 ,
(1)
where ρm is the resistivity of the pure aluminum matrix and is a constant that depends strongly on temperature. csolut, cvac, cdisl, cGB, and cprecip are the solute concentration, vacancy concentration, dislocation density, grain boundary fraction, and density of precipitates in the matrix, respectively, with the constants ρsolut, ρvac, ρdisl, ρGB, and ρpreci accounting for the contribution of each of these factors to the resistivity in the aluminum, respectively. In addition, i and j represent solute elements and precipitates, respectively (e.g., i=Mg, Zn, Cu and j=GP, η′, η). In the present case, the contribution of precipitates to the resistivity was estimated, and the other influencing factors were neglected. On the basis of research on the effect of a single precipitate on the resistivity by Raeisinia et al. [18], we used similar estimates for the resistivity: 𝜌𝑝𝑟𝑒𝑐𝑖𝑝 = 𝑋/√𝑠 ,
(2)
where s is the spacing of precipitates (in nm) and X is a constant(X=12 for 6000 series Al alloys). It is evident that the contribution of precipitates to the electrical resistivity calculated using equation (3) and Table 2 follows the order of ρprecip(500°C) > ρprecip(300°C) > ρprecip(400°C) (In this study, only qualitative analysis be adopted), which corresponds to the electrical conductivity following the order of κprecip(400°C) > κprecip(300°C) > κprecip(500°C). This estimate is in relatively good agreement with the electrical conductivity in a stable state from Fig. 8(c). For the hardness, each microstructural feature has the same effect. Therefore, the hardness can be estimated using Eq. (3): 𝑖 𝐻𝑉 = 𝐻𝑉𝑚 + ∑𝑖 ∆𝐻𝑉𝑠𝑜𝑙𝑢𝑡 + ∆𝐻𝑉𝑣𝑎𝑐 + ∆𝐻𝑉𝑑𝑖𝑠 + ∆𝐻𝑉𝐺𝐵 + ∆𝐻𝑉𝑝𝑟𝑒𝑐𝑖𝑝 ,
(3)
where HVm is the hardness of the pure Al matrix, and HVsolut, HVvac, HVGB, HVprecip are the hardening effect originating from the vacancies, dislocations, grain boundaries, solute elements, and precipitates in the aluminum matrix, respectively. For the contribution of precipitates to the hardness, the approach proposed by Chen et al. [19] was followed, assuming that the dislocations bypass precipitates using the Orowan mechanism. Then, the corresponding hardness increase can be written as ∆𝐻𝑉𝑝𝑟𝑒𝑐𝑖𝑝 = 0.81𝐶𝑀𝐺𝑏𝑙𝑛(𝑑𝑚𝑒𝑎𝑛 /𝑏)/[2𝜋𝑑𝑚𝑒𝑎𝑛 (√1 − υ )(0.615√2𝜋/3𝑉𝑓 − 1)],
(4)
where C is a constant (C=3.06～3.28 for Al alloys[20,21], and Cmean=3.16 is adopted); M is the mean orientation factor (M=3.06 for the face centered cubic (f.c.c) Al matrix)[20,22];G is the shear modulus (G=27 GPa for 7000 series Al alloys)[23,24]; b is the Burgers vector (b= 0.286 nm for the f.c.c Al)[23,24]; υ is the Poisson’s ratio (υ=0.33 for the Al)[25]; Vf and dmean are respectively the volume fraction and the mean diameter of precipitates, which are listed in Table 3. The results were obtained using equation (3) and Table 3 and followed the order of ΔHVprecip(300°C) > ΔHVprecip(400°C) > ΔHVprecip(500°C) (ΔHVprecip(300°C)=23.041HV, ΔHVprecip(400°C)=22.682HV, ΔHVprecip(500°C)=21.234HV). Thus, the results of this approach correspond well with the measured hardness in a stable state (Fig. 8(a)), which further supports the present model. In a complex alloy system, multiple strengthening mechanisms are effective, including grain boundary strengthening, dislocation strengthening, precipitation strengthening and solid solution strengthening. A recent study revealed that for a 7075 Al alloy, grain boundary strengthening is the dominant strengthening mechanism[26]. For the 7N01 alloy, although the original material was rolled, the material did not recrystallize due to the short holding time during thermal cycling, taking into account that the grain size of the material during natural aging kept invariant. Thus, it is reasonable to state that the grain boundary strengthening is a relatively small contributor to the total strength variation in this study. Dislocation 8
strengthening is mainly affected by the processing deformation degree of the alloy. For the same processing state materials, the contribution of dislocation strengthening to the total strength could be considered as same. In the present study, it is comfirmed that both precipitation strengthening and solid solution strengthening are dominant mechanisms for 7N01 during natural aging after thermal cycle. According to previous research reported by KaKa[26], the solid solution strengthening is a relatively small contributor compare to the precipitation strengthening in 7000 series alloy. Moreover, the precipitation would reduce during the later natural aging. For the contribution of precipitation strengthening to the yield stress, following the Orowan equation [27]: 𝑝𝑟𝑒𝑐𝑖𝑝
∆𝜎0.2
= 0.4𝑀𝐺𝑏 ln(2𝑟𝑚𝑒𝑎𝑛 /𝑏) /(𝜋𝑠√1 − 𝑣 ).
(5) 𝑝𝑟𝑒𝑐𝑖𝑝
Likewise, the yield strength increments were arranged in the sequence of ∆𝜎0.2 𝑝𝑟𝑒𝑐𝑖𝑝
∆𝜎0.2
(300℃)＞
𝑝𝑟𝑒𝑐𝑖𝑝 (400℃)＞∆𝜎0.2 (500℃), which is consistent with the tendency of the curve in Fig. 8(b).
3.4 Microstructure of precipitation observed by TEM analysis
Fig. 9. TEM images showing distribution of strengthening precipitates at different magnifications: a) 500°C 9
and b) and c) 400°C As evidenced by the TEM images in Fig. 9, the strengthening mechanism for 7N01 is precipitation strengthening. Dislocation movement is controlled by the distribution of these precipitates, which act as pinning points (Fig. 9(a) and (b)). Precipitates exhibiting a symmetrical crescent shape are observed in Fig. 9(c) , which appear to be cut from the center line. According to previous research reported by Wang [28], this feature indicates that the dislocation encountered two barriers from shearable and non-shearable particles during movement. Finally, as the size, volume fraction, and spacing of precipitates have significant effects on the mechanical and electrical properties of a material, these properties will be discussed in the following paragraphs. Fig. 10 is the HRTEM images of the precipitates enclosed by the red circle in Fig. 9(c). The precipitate was identified as hexagonal η (MgZn2) phase with an orientation relationship to the Al matrix of (1120)η ∥(220)Al and (0002)η∥(002)Al. Fig. 10. (c) and (e) are the atomic model and simulated diffraction pattern of MgZn2 viewed from [1100]η, respectively, which has a hexagonal lattice with a=5.223Å and c=8.556Å (crystal data proposed by Wandahl et al. [29]). Obviously, It is in relatively good agreement between the observed structure (Fig. 10 (b) and (d)) and the atomic model (Fig. 10 (c) and (e)). Therefore, this precipitate structure can be identified as MgZn2. In addition, the Inverse fast Fourier transformation (IFFT) images in Fig. 10.(f) and (g), which were obtained from the (220) Al/ (2 2 0) Al and(002)Al/(002)Al diffraction patterns, respectively, clearly shows the existence of the dislocation along the interface between the η precipitate and the Al matrix. The lattice misfit between (1120)η and (220)Al is 45.3% ( d (1120) η = 0.2611nm, d (220) Al = 0.1427 nm, calculated as δ =(d (1120) η － d(220)Al)/d(1120)η). Hence, the interface between particle and matrix was non-coherent.
Fig. 10. (a) HRTEM images of the η(MgZn2) precipitates viewed from [110]Al; (b) the IFFT pattern after array mask of the framed area in the HREM image of (a); (c) atomic model of MgZn2 viewed from [1100]η;(d) and (e) the FFT pattern of the region discussed and simulated diffraction pattern of MgZn2 10
viewed from [1100]η;(f) and (g) the IFFT pattern obtained using the (220)Al/(2 2 0)Al and(002)Al/(002)Al diffraction patterns of the region discussed.
Fig. 11. Bright-field TEM micrographs for different peak temperatures after natural aging for 10 weeks: a) 300°C, b) 400°C, and c) 500°C The bright-field TEM micrographs in Fig. 11 show the microstructural evolution of the natural aged 7N01 alloys after the welding thermal cycles. The microstructure of the alloy for the peak temperature of 300°C consisted of coarse precipitates in which η′ precipitates were homogeneously distributed (Fig. 11(a)). Fig. 11(b) shows that the distribution of the precipitates for the alloy with the peak temperature of 400°C was similar to that for the peak temperature of 300°C. However, a significant decrease in the particle size was observed for the peak temperature of 500°C. This distribution could result from mostly dissolution of η′ and η above 400°C during the thermal cycle and dynamic precipitation process (the cooling stage of the thermal cycle and subsequent natural aging period), leading to the nucleation and growth of GP zones, which can be corroborated by the schematic in Fig. 7. In combination with the volume fraction, Eq. (4) was used to estimate the contribution of the 11
precipitates to the hardness, and then, ΔHVprecip was calculated. The volume fraction of the precipitates outlined by Pellissier and Purdy [30] can expressed as: 2 𝑉𝑓 = 1.4𝜋𝑁𝑑𝑚𝑒𝑎𝑛 /(6𝐴),
(6)
where N is the number of precipitates in the observation area and A is the image area. For the spacing between precipitates, it was hypothesized that the effect of precipitates on the mechanical and electrical properties of the Al alloy should scale as the spacing between precipitates. Then, according to the work of Dixit et al. [31], the corresponding spacing can be written as 𝑠 = 𝑟𝑚𝑒𝑎𝑛 (√2𝜋/3𝑉𝑓 − √8/3 ).
(7)
Table 3 Statistical results of intragranular precipitates for different peak temperatures of thermal cycle Peak temperature(℃) 300 400 500 6 2 Image area(*10 nm ) 7.040 19.526 11.807 Number of particles 54±9 128±15 158±30 Mean particle radius(nm) 60.705±3.21 69.885±2.32 25.345±6.13 Area fraction of particle (%) 8.647±1.13 7.593±0.13 2.952±0.43 Volume fraction of particle (%) 8.166±0.509 9.359±0.512 2.436±0.711 Spacing(nm) 209.257±20.626 217.147±16.243 192.868±12.473
Fig. 12. Particle size distribution (measured from TEM data)for different peak temperatures of thermal cycle: a) 300°C, b) 400°C, and c) 500°C 12
The precipitate size and distribution evolution for different processing temperatures were evaluated from bright-field TEM images. Part of the dislocations around the precipitates may affect the contrast and make TEM observations difficult, leading to large statistical results. To allow for a quantitative bulk measurement of the mean size and fraction of precipitates, it was necessary to count at least two bright-field TEM images containing less dislocations. In addition, because an increase in the sample size can assure the randomness of the measured sample and thus result in a reasonable overall distribution, the counting was performed using low-magnification TEM images. The effect of peak temperature on strengthening precipitates was determined by measuring the area fraction of particles, which refers to the percentage of the image area occupied by particles. The calculated area fractions of particles are listed in Table 3, revealing a gradual decrease during the welding thermal cycle with increasing peak temperature from 300°C to 500°C. It is evident from Fig. 12 that the particle size followed an approximately normal distribution. Moreover, the peak frequency of the particle radius did not vary greatly with temperature from 300°C to 400°C; however, the particle radius was small at 500°C (r300=50–70 nm, r400=50–70 nm, r500=10–20 nm). In addition, compared with the results for the peak temperatures of 300°C and 400°C, the mean value was significantly reduced for 500°C, as summarized in Table 3. Notably, a high frequency of particle sizes of r300=10–20 nm and r400＞100 nm are observed in Fig. 12(a) and (b), respectively. This finding indicates that the cluster and GP zone formation, using atoms provided by the dissolution of GP zones and η′ during the welding thermal cycle, was promoted during natural aging with the 300°C peak temperature and that coarsening of η occurred during the welding thermal cycle for the peak temperature of 400°C (coarsening of η consumes atoms used to form the GP zones during natural aging). This estimate is in good agreement with the schematic in Fig. 7. For the peak temperature of 500°C, the average size, area fraction, and peak frequency of precipitates were observed to decrease significantly. These results may be attributable to the (1) dissolution temperature of original large precipitates being reached during the welding thermal cycle, (2) destruction of precipitates by dislocation movement, and (3) formation of new small precipitates during the natural aging period. From the previous discussion (Fig. 7), the solvus temperature of η′ is below 300°C, thereby supporting option (1). The TEM observations in Fig. 9(c) revealed the symmetrical crescent shape of the particles, which is evidence of dislocation cutting. The dislocations passed through the second-phase particles and only slipped when the external stress exceeded the internal stress of the second-phase particles. The specimen in this experiment was only affected by the heat effect during processing and not by plastic deformation. Because the cutting was not caused by thermal cycling and the original material was subjected to deformation by rolling, option (2) can be eliminated. In addition, the peak frequency of the particle radius for the peak temperature of 500°C was r500=10–20 nm, which indicates that new precipitates originated from the GP zones during cooling of the thermal cycling and natural aging period, thereby validating the third option. Hence, it can be concluded that the severe deterioration of the properties of the evaluated particles with progressing peak temperature can be mainly attributed to precipitate dissolution. 4. Conclusion The microstructural evolution and mechanical behavior of a 7xxx alloy after welding thermal cycling and late natural aging were investigated. The effect of different peak temperatures of the welding thermal cycle on the precipitates and physical performance of Al alloys was analyzed, and the following conclusions were drawn: (1) High heating rates lead to the hysteresis effect of the reaction during DSC scanning; thus, the solvus temperature range of precipitates of 7N01 alloy during welding thermal cycling is higher than that measured using DSC. The solvus temperature of η′ was determined to be between 200°C and 300°C during the 13
welding process. Moreover, the formation and coarsening temperature of η was close to 400°C and the solvus temperature of η was between 400°C and 500°C during the welding process. (2) The Al–Zn–Mg–Cu alloy processed by thermal welding cycle exhibited obvious softening behavior; however, the hardness and strength were improved during natural aging. In addition, electrical conductivity measurements indicated that the SCC resistance was enhanced during the natural aging period. (3) Quantitative experimental data obtained from bright-field TEM images of the Al–Zn–Mg–Cu alloy revealed good consistency between the statistical indicators and physical properties during the natural aging period. Therefore, this work indicates that this reasonable statistic methodology could be introduced to predict the changes of the physical properties of these materials. Acknowledgements The present study was supported by the National Key Research and Development Program of China (Grant No. 2016YFB0700505). The authors thank Dr. D. C. Zhu and Mr. R. J. Huang from the Materials College of Sichuan University for their help with the electrical conductivity tests. References [1] Sachdev, A. K., Mishra, R. K., Mahato, A., & Alpas, A. (2012). Vehicle Lightweighting: Challenges and Opportunities with Aluminum. ICAA13: 13th International Conference on Aluminum Alloys. John Wiley & Sons, Inc. [2] Grong, Ø. (1997). Metallurgical modelling of welding. Institute of Materials. [3] B. I. Bjørneklett, Ø. Grong, O. R. Myhr, & A. O. Kluken. (1999). A process model for the heat-affected zone microstructure evolution in al-zn-mg weldments. Metallurgical & Materials Transactions A, 30(10), 2667-2677. [4] Sato, Y. S., Kokawa, H., Enomoto, M., & Jogan, S. (1999). Microstructural evolution of 6063 aluminum during friction-stir welding. Metallurgical & Materials Transactions A, 30(9), 2429-2437. [5] Hansen, V., Karlsen, O. B., Langsrud, Y., & Gjønnes, J. (2013). Precipitates, zones and transitions during aging of al-zn-mg-zr 7000 series alloy. Materials Science & Technology, 20(2), 185-193. [6] Zhang, S., Hu, W., Berghammer, R., & Gottstein, G. (2010). Microstructure evolution and deformation behavior of ultrafine-grained al–zn–mg alloys with fine η ′ precipitates. Acta Materialia, 58(20), 6695-6705. [7] Deschamps, A., Fribourg, G., Bréchet, Y., Chemin, J. L., & Hutchinson, C. R. (2012). In situ evaluation of dynamic precipitation during plastic straining of an al–zn–mg–cu alloy. Acta Materialia, 60(5), 1905-1916. [8] Deschamps, A., & Geuser, F. D. (2013). Quantitative characterization of precipitate microstructures in metallic alloys using small-angle scattering. Metallurgical & Materials Transactions A, 44(1), 77-86. [9] Dumont, M., Lefebvre, W., Doisneau-Cottignies, B., & Deschamps, A. (2005). Characterisation of the composition and volume fraction of η ′ and η, precipitates in an al–zn–mg alloy by a combination of atom probe, small-angle x-ray scattering and transmission electron microscopy. Acta Materialia, 53(10), 2881-2892. [10] Hutchinson, C. R., Geuser, F. D., Chen, Y., & Deschamps, A. (2014). Quantitative measurements of dynamic precipitation during fatigue of an al–zn–mg–(cu) alloy using small-angle x-ray scattering. Acta Materialia, 74(74), 96-109. [11] Liang, Z., Li, X., Nie, Z., Hui, H., Sun, J., & Sun, Z. (2016). Microstructure and mechanical properties of joints of a new al-zn-mg-cu alloy welded by tig. Rare Metal Materials & Engineering. 14
[12] Deschamps, A., Livet, F., & Bréchet, Y. (1998). Influence of predeformation on ageing in an al–zn–mg alloy—i. microstructure evolution and mechanical properties. Acta Materialia, 47(1), 281-292. [13] Buha, J., Lumley, R. N., & Crosky, A. G. (2008). Secondary ageing in an aluminium alloy 7050. Materials Science & Engineering A, 492(1–2), 1-10. [14] Jr, A. F. O., Barros, M. C. D., Cardoso, K. R., & Travessa, D. N. (2004). The effect of rra on the strength and scc resistance on aa7050 and aa7150 aluminium alloys. Materials Science & Engineering A, 379(1–2), 321-326. [15] Park, J. K., & Ardell, A. J. (1991). Microchemical analysis of precipitate free zones in 7075-a1 in the t6, t7 and rra tempers. Acta Metallurgica Et Materialia, 39(4), 591-598. [16] Li, X. M., & Starink, M. J. (2000). Analysis of Precipitation and Dissolution in Overaged 7xxx Aluminium Alloys using DSC. Materials Science Forum (Vol.331-337, pp.1071-1076). [17] Pouraliakbar, H., Jandaghi, M. R., & Khalaj, G. (2017). Constrained groove pressing and subsequent annealing of al-mn-si alloy: microstructure evolutions, crystallographic transformations, mechanical properties, electrical conductivity and corrosion resistance. Materials & Design, 124, 34-46. [18] Raeisinia, B., Poole, W. J., & Lloyd, D. J. (2006). Examination of precipitation in the aluminum alloy aa6111 using electrical resistivity measurements. Materials Science & Engineering A, 420(1–2), 245-249. [19] Chen, Y., Gao, N., Sha, G., Ringer, S. P., & Starink, M. J. (2016). Microstructural evolution, strengthening and thermal stability of an ultrafine-grained al–cu–mg alloy. Acta Materialia, 109, 202-212. [20] Xiao Guang Qiao, Nong Gao, & Marco J. Starink. (2012). A model of grain refinement and strengthening of al alloys due to cold severe plastic deformation. Philosophical Magazine, 92(4), 446-470. [21] Qiao, X. G., Starink, M. J., & Gao, N. (2009). Hardness inhomogeneity and local strengthening mechanisms of an al1050 aluminium alloy after one pass of equal channel angular pressing. Materials Science & Engineering A, 513(11), 52-58. [22] Taylor, G. I. (1934). The mechanism of plastic deformation of crystals. part i. Theoretical. Proceedings of the Royal Society of London, 145, 362-387. [23] Ma, K., Wen, H., Hu, T., Topping, T. D., Isheim, D., & Seidman, D. N., et al. (2014). Mechanical behavior and strengthening mechanisms in ultrafine grain precipitation-strengthened aluminum alloy. Acta Materialia, 62(5), 141-155. [24] Chen, Y., Gao, N., Sha, G., Ringer, S. P., & Starink, M. J. (2015). Strengthening of an al-cu-mg alloy processed by high-pressure torsion due to clusters, defects and defect-cluster complexes. Materials Science & Engineering A, 627, 10-20. [25] Wang, S. C., Lefebvre, F., Yan, J. L., Sinclair, I., & Starink, M. J. (2006). Vppa welds of al-2024 alloys: analysis and modelling of local microstructure and strength. Materials Science & Engineering A, 431(1–2), 123-136. [26] Ma, K., Hu, T., Yang, H., Topping, T., Yousefiani, A., & Lavernia, E. J., et al. (2016). Coupling of dislocations and precipitates: impact on the mechanical behavior of ultrafine grained al–zn–mg alloys. Acta Materialia, 103(15), 153-164. [27] Wen, H., Topping, T. D., Isheim, D., Seidman, D. N., & Lavernia, E. J. (2013). Strengthening mechanisms in a high-strength bulk nanostructured cu–zn–al alloy processed via cryomilling and spark plasma sintering ☆. Acta Materialia, 61(8), 2769-2782. [28] Wang, X. (2005). The shearable–non-shearable transition in al–mg–si–cu precipitation hardening alloys: implications on the distribution of slip, work hardening and fracture. Philosophical 15
Magazine, 85(26-27), 3113-3135. [29] Wandahl, G., Christensen, A. N., Uggerud, E., Songstad, J., Lönnberg, H., & Colacio, E., et al. (1989). Occurrence of extinction correlated with crystal growth mode for crystals of pd3ce and mgzn2. Acta Chemica Scandinavica, 43(3), 296-297. [30] Pellissier, G. E., Purdy, S. M. (1972). Stereology and Quantitative Metallography. American Society for Testing & Materials. [31] Dixit, M., Mishra, R. S., & Sankaran, K. K. (2008). Structure–property correlations in al 7050 and al 7055 high-strength aluminum alloys. Materials Science & Engineering A, 478(1–2), 163-172.
16
www.elsevier.com/locate/msea
PII: DOI: Reference:
S0921-5093(18)30096-0 https://doi.org/10.1016/j.msea.2018.01.067 MSA36021
To appear in: Materials Science & Engineering A Received date: 23 October 2017 Revised date: 17 January 2018 Accepted date: 18 January 2018 Cite this article as: Kang Zhang, J.Q. Chen, P.Z. Ma and X.H. Zhang, Effect of welding thermal cycle on microstructural evolution of Al–Zn–Mg–Cu alloy, Materials Science & Engineering A, https://doi.org/10.1016/j.msea.2018.01.067 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 welding thermal cycle on microstructural evolution of Al–Zn–Mg–Cu alloy Kang Zhanga, J. Q. Chena,*, P. Z. Maa, X. H. Zhanga a Key Laboratory of Advanced Technologies of Materials, Ministry of Education, School of Materials Science and Engineering, Southwest Jiaotong University, Chengdu 610031, China * [email protected] ABSTRACT Al–Zn–Mg–Cu alloys are used extensively in high-speed train applications; however, the occurrence of softening in the heat-affected zone after welding limits their development. In this work, the effect of the welding thermal cycle on the softening and aging behaviors of a 7xxx aluminum alloy was investigated by examining the state of the strengthening precipitates. The microstructure and solvus temperature range of the precipitates of 7N01P-T4 were characterized using TEM and DSC analysis, respectively. It was found that the softening behavior of the aluminum alloys was closely related to the volume fraction and size of the hardening precipitates, which were greatly affected by the peak temperature of the welding thermal cycle. In addition, η′ and η precipitates were observed to be primarily responsible for the increase in the mechanical and electrical properties during the room-temperature natural aging. A phenomenological connection was thus uncovered between the characteristic parameters of the thermal cycle and the precipitation behavior, providing insight for the design of the welding process for 7N01 alloys. Keywords: Al–Zn–Mg–Cu alloy, thermal cycle, softening behavior, TEM, precipitation 1. Introduction With the trend of the development of light-weight structures, high-strength aluminum alloys have been used extensively. Alloy 7N01, a heat-treatable Al–Mg–Zn alloy, was developed for train body applications in the high-speed train industry [1]. In these applications, the alloys must be joined using a welding process; therefore, the effect of welding on the material properties is an important consideration. Unfortunately, softening generally occurs in the heat-affected zone (HAZ) of heat-treatment-strengthened Al alloys during the welding process. In addition, the location of the softened zone is typically a potential fatigue failure location during in-service processes, which seriously affects the safety of the structure, reported by Grong [2]. Therefore, in the present research, the coupling of dissolution and precipitation in the HAZ and the peak temperature in the thermal cycle was investigated, as it is important for clarifying the softening problem in the 7N01 welding process. Unlike heat treatment, a welding thermal cycle is a rapid heating and cooling process. The dissolution and precipitation of strengthening phases in the HAZ occur during the welding cycle and subsequent natural aging. Bjørneklett et al. [3] used process modeling techniques to develop a model that provides a sound basis for predicting the HAZ strength loss during welding of Al–Zn–Mg alloys as well as the resulting recovery of strength after prolonged natural aging. Because the HAZ of the weld was too narrow to prepare test specimens, Sato et al. [4] performed thermal simulation to evaluate specimens with a wide uniform temperature zone, and subsequent transmission electron microscopy (TEM) observation revealed that the precipitate distribution corresponded to the hardness profile in the HAZ of Al alloys. Hansen et al. [5] evaluated the effect of time–temperature ranges of the precipitate transformations of 7000 series alloys during the aging process. In addition, the dependence on the quenching temperature was analyzed using differential scanning calorimetry (DSC) and TEM. The mechanical properties of precipitation-strengthened Al alloys were observed to be closely related to the volume fraction and size of the hardening precipitates. Zhang et al. [6] monitored the evolution of the 1
precipitate size with progressing deformation using bright-field TEM images. The statistical results revealed an association between the microstructural evolution and mechanical behavior of an aged 6xxx alloy with finely dispersed η′ precipitates. In addition, small-angle X-ray scattering (SAXS) was recently used for an in situ quantitative study at elevated temperature of the mean size and fraction of precipitates in a 7xxx alloy [7,8,9], Hutchinson et al. [10] used SAXS for an in situ, elevated temperature study of the effect of monotonic strain and strain rate on the precipitate distribution in a 7xxx alloy，their results show that SAXS is mostly sensitive to the measurement of the precipitation state. However, because of the limitation of the heating mode and synchrotron radiation, studies of the evolution of precipitates during the welding of aluminum alloys using SAXS have seldom been reported. These considerations point to the need for a statistical methodology to obtain an in-depth understanding of precipitates, namely a quantitative measure of the mean size and fraction of precipitates for investigation of the softening problem for Al–Zn–Mg–Cu alloys. DSC and TEM are conventional methods that meet this requirement. Characterization of the solvus temperature range and mean radii of precipitates by DSC and TEM, respectively, can be used to examine the relationship between strengthening precipitates and thermal cycles. However, studying thermal cycle–precipitation coupling using this method slightly reduces the precision of the observations, partly because TEM itself can only be used to examine the micro-regional regularity of small precipitates and because the dislocations around precipitates make TEM observations difficult. The purpose of this study was not only to reveal the relationship between hardened precipitates and the thermal cycle but also to complement studies that have previously examined whether the physical performance of 7N01 alloy could be improved during the natural aging period. In this case, the use of TEM and DSC was acceptable, and the HAZ samples were processed using the thermal simulation technique. Hence, the specific purpose of this work was to investigate (i) the effect of thermal cycling on precipitates of a 7xxx alloy and (ii) the evolution of physical properties of materials (after being processed by thermal cycling) during a natural aging period. The generally poor mechanical performance of the 7xxx Al alloys was evaluated directly after the thermal cycle, and precipitate strengthening was considered after subsequent natural aging. This study of the HAZ softening problem will provide insight for the design of a more reasonable welding process. 2. Materials and experimental preparation 2.1 Temperature measurement during welding process In this study, the peak temperature of the thermal cycle in the HAZ of 7N01P-T4 aluminum alloy was first characterized. The chemical composition of 7N01P-T4 is provided in Table 1. Table 1 Chemical composition of aluminum alloy 7N01P-T4 (wt. %) material
Si
Fe
Cu
Mn
Mg
Cr
Ni
Zn
Ti
V
Zr
Al
7N01P-T4
＜0.001
0.087
0.032
0.377
1.049
0.089
＜0.001
4.23
0.029
0.014
0.083
Bal.
The HAZ of the 7N01 alloy was subjected to thermal cycling using one-pass hybrid laser–metal inert gas (MIG) welding with the welding parameters listed in Table 2. During the welding process, a pair of K-type thermocouples was used to monitor the temperature evolution of the thermal cycle. The thermocouples were arranged as shown in Fig. 1 at different distances (0, 3, 5, 7, 10, and 15 mm) from fusion line in the transverse direction. Table 2 Parameters of hybrid laser–MIG welding Laser power Wire speed Speed welding Welding voltage Welding current (Kw) (m/min) (m/min) (V) (A) 2
2
9
0.72
22.4
240
Fig. 1. Schematic diagram of welded plate used for measurement of thermal cycle showing the layout of the thermocouples 2.2 Mechanical property measurement
Fig. 2. Schematic diagram of specimen for thermal cycle simulation (specimens were cut with the longitudinally along the rolling direction) (a) tensile test specimen, (b) hardness and electrical conductivity specimen Firstly, the simulated thermal cycles (acquired by thermocouples) were loaded on the specimens by Gleeble machine. After being processed using the different thermal cycles, the Vickers hardness (3 kg load held for 10 s), electrical conductivity (%IACS)( as a supplement to the hardness to characterize the strengthening phase changes after thermal cycle) and tensile test (strain rate of 3mm/min) were measured once a week to monitor the softening behavior of the alloy with natural aging. The test specimen for the thermal–mechanical simulator is shown as Fig. 2. The hardness, electrical conductivity and tensile test were performed with a HVS-30 tester, a 7501 eddy current conductivity meter and a Gleeble machine, respectively. The mean value of ten measurement data was taken as the hardness and the electrical conductivity of the corresponding condition. 2.3 DSC and TEM observations DSC measurements were performed using a TA Instruments Discovery DSC calorimeter at heating 3
rates of 5°C/min, 25°C/min, and 50°C/min, respectively. High-purity aluminum of the same weight was used as a reference. During the test, the chamber was evacuated and then backfilled with pure argon gas. The microstructural changes occurring in the softening zone with the thermal cycling were examined using TEM. Thin foils specimens 30μm in thickness were obtain from the plates (Fig.2. (b)) for TEM observation. Electron-transparent sections of thin foils were prepared using a twin-jet electropolisher with a mixture of 70% methanol and 30% nitric acid (−25°C with liquid nitrogen). An FEI Titan G2 60-300 microscope operating at 200 kV was used for the TEM observation. The crystal models were built by Crystal Maker software. The measurements of the TEM images were done with the Image J software and Digital Micrograph software. 3. Results and discussion 3.1 Softening behavior after welding thermal cycles According to previous research [11], the HAZ of 7xxx alloys consists of two parts: a solid-solution zone and an overaging zone. The width of the HAZ in one-pass welding of 7N01 is approximately 3–10 mm, which corresponds to peak temperatures of approximately 150°C–550°C, as shown in Fig. 3, which were used to the primary peak temperatures of the simulated thermal cycle. First, thermal cycles at different peak temperatures of 150°C, 200°C,300°C, 400°C, and 500°C were simulated using a Gleeble-3500 thermal–mechanical simulator. Corresponding to the temperarure measurement during welding process, the specimen was heated to the peak temperature at a rate of 150°C/s and held at the peak temperature for 2 s, then air cooled to room temperature, as illustrated in Fig. 4. Tensile tests at room temperature were performed immediately after the samples were subjected to welding thermal cycles. Figure 5 reveals the severe decline in strength of the aluminum alloys after being subjected to thermal cycles with peak temperatures between 300°C and 500°C. (initial.T4 in Fig.5 stands for the initial state of the material without being subjected to thermal cycle ). This finding indicates that a large number of precipitates dissolved in this temperature range. Thus, peak temperatures of 300°C, 400°C, and 500°C were used to study the mechanical properties of the samples during the late natural aging period.
Fig. 3. Peak temperature distribution of 7N01 alloy measured by thermocouple during welding experiment
4
Fig. 4. The thermal cycles applied to 7N01 alloy
Fig. 5. Stress–strain curves of 7N01 alloy after being subjected to welding thermal cycles with different peak temperatures by the thermal–mechanical simulator 3.2 DSC analysis of the precipitate transition of 7N01P-T4
5
Fig. 6. DSC curves of 7N01P-T4 at different heating rates To better understand the specific relationship between the thermal cycle and precipitates, DSC tests were conducted. The effect of the heating rate on the initial state of 7N01P-T4 alloy is shown in Fig. 6. The number of peaks in the DSC curve is related to the nature of the material. The curves consist of wide and sharp peaks (endothermic and exothermic), with the location of the peaks changing depending on the heating rate. The differences in the peaks in Fig. 6 are considered to be mainly determined by the equilibrium phase transition suppressed , which causes the hysteresis effect of reaction. Fig. 6 further reveals that the heat effect of precipitate dissolution, formation, and growth depends on the heating rate. According to previous research reported by Deschamps et al. [12], the first peak of the DSC curve corresponds to GP zone formation, which generally occurs below 150°C in Al–Zn–Mg–Cu alloys; however, no significant endothermic or exothermic peak due to GP zone formation or dissolution was observed in the present alloys. This finding may indicate that the cluster and GP zone formation was completed in the present state of the alloy. As observed in Fig. 6, Peak I (the first endothermic peak) is a combination of two main reactions, mostly the dissolution of η′ and fine η. Peak II (the first exothermic peak) represents the formation and coarsening of η and corresponds to the determination of the activation energy of η. The amounts of η′ and η in the alloy depend on the Mg and Zn contents in 7N01, which, as a Zn-rich Al–Zn–Mg alloy, has a greater tendency to form precipitates described in Buha et al. [13]. Peak III (the second endothermic peak) corresponds to the dissolution of η. Because of the hysteresis effect caused by the high heating rate, the transformation temperature of the endothermic peak is higher, thereby delaying the corresponding dissolution effect during the welding thermal cycles in the HAZ. Combined with the observations in Fig. 5, a serious decline in strength occurred after the thermal cycles with peak temperatures between 300°C and 550°C, as determined by the dissolution of η′ and the formation/coarsening of η. Hence, Fig. 7 shows the dissolution and reprecipitation sequence of precipitates during the welding thermal cycles in the HAZ. In the initial T4 state of 7N01, precipitates are homogeneously distributed with sufficient growth. During the welding thermal cycle, dissolution of GP zones, η′, and fine η occurred during the heating period, and the dissolution of solute atoms back into the Al matrix occurred during the cooling period.
Fig. 7. Schematic diagram of dissolution during welding thermal cycle 6
3.3 Mechanical property of 7N01P-T4 alloy during the following natural aging
Fig. 8. Effect of natural aging and peak temperature of the thermal cycle on the (a) vicker’s hardness; (b) 0.2% proof strength of 7N01 aluminum alloy; (c) electrical conductivity. In the present study, the effect of the peak temperature of the thermal cycle on the mechanical and electrical properties was investigated during natural aging. To evaluate the mechanical properties of 7N01 aluminum alloy during natural aging, hardness and tensile tests were performed. With heat treatment condition change, it exhibits a correlation between the conductivity and mechanical properties, which could reflect the state of strengthening phase. Therefore, the electrical conductivity was measured to evaluate the transformation of precipitation and stress corrosion cracking (SCC) resistance of the alloy during natural aging[14,15,16]. As observed in Fig. 8, the hardness, strength and electrical conductivity of the 7N01 alloy after the welding thermal cycle are plotted as a function of the natural aging time. Increasing the natural aging time clearly led to a decrease of the electrical conductivity and increase of the hardness and strength. Moreover, it was concluded that a higher peak temperature leads to a lower electrical conductivity and hardness when the curve tends to be stable. A similar trend was observed for the strength. One week after the welding thermal cycle, significant softening to the lowest hardness value was observed, with a value significantly lower than that of the initial T4 state (124HV). Similarly, the strength was lower than that of the the initial T4 state (313 MPa). In contrast, the conductivity increased significantly (30%IACS of the T4 state), which was caused by the dissolution of precipitates and the re-dissolution of a large number of solute atoms into the Al matrix. The Matthiessen–Fleming rule is typically used to describe the electrical resistivity (the reciprocal of electrical conductivity) of materials. However, the electrical resistivity of multiphase alloy is relatively complex because of the effect of multiple precipitates, and no consensus regarding its behavior has been reached. A simple and modified Matthiessen rule can be used to describe the effect of precipitates on 7
electrical resistivity [17]: 𝑗
𝑖 𝑖 𝜌 = 𝜌𝑚 + ∑𝑖 𝑐𝑠𝑜𝑙𝑢𝑡 𝜌𝑠𝑜𝑙𝑢𝑡 + 𝑐𝑣𝑎𝑐 𝜌𝑣𝑎𝑐 + 𝑐𝑑𝑖𝑠 𝜌𝑑𝑖𝑠 + 𝑐𝐺𝐵 𝜌𝐺𝐵 + ∑𝑗 𝜌𝑝𝑟𝑒𝑐𝑖𝑝 ,
(1)
where ρm is the resistivity of the pure aluminum matrix and is a constant that depends strongly on temperature. csolut, cvac, cdisl, cGB, and cprecip are the solute concentration, vacancy concentration, dislocation density, grain boundary fraction, and density of precipitates in the matrix, respectively, with the constants ρsolut, ρvac, ρdisl, ρGB, and ρpreci accounting for the contribution of each of these factors to the resistivity in the aluminum, respectively. In addition, i and j represent solute elements and precipitates, respectively (e.g., i=Mg, Zn, Cu and j=GP, η′, η). In the present case, the contribution of precipitates to the resistivity was estimated, and the other influencing factors were neglected. On the basis of research on the effect of a single precipitate on the resistivity by Raeisinia et al. [18], we used similar estimates for the resistivity: 𝜌𝑝𝑟𝑒𝑐𝑖𝑝 = 𝑋/√𝑠 ,
(2)
where s is the spacing of precipitates (in nm) and X is a constant(X=12 for 6000 series Al alloys). It is evident that the contribution of precipitates to the electrical resistivity calculated using equation (3) and Table 2 follows the order of ρprecip(500°C) > ρprecip(300°C) > ρprecip(400°C) (In this study, only qualitative analysis be adopted), which corresponds to the electrical conductivity following the order of κprecip(400°C) > κprecip(300°C) > κprecip(500°C). This estimate is in relatively good agreement with the electrical conductivity in a stable state from Fig. 8(c). For the hardness, each microstructural feature has the same effect. Therefore, the hardness can be estimated using Eq. (3): 𝑖 𝐻𝑉 = 𝐻𝑉𝑚 + ∑𝑖 ∆𝐻𝑉𝑠𝑜𝑙𝑢𝑡 + ∆𝐻𝑉𝑣𝑎𝑐 + ∆𝐻𝑉𝑑𝑖𝑠 + ∆𝐻𝑉𝐺𝐵 + ∆𝐻𝑉𝑝𝑟𝑒𝑐𝑖𝑝 ,
(3)
where HVm is the hardness of the pure Al matrix, and HVsolut, HVvac, HVGB, HVprecip are the hardening effect originating from the vacancies, dislocations, grain boundaries, solute elements, and precipitates in the aluminum matrix, respectively. For the contribution of precipitates to the hardness, the approach proposed by Chen et al. [19] was followed, assuming that the dislocations bypass precipitates using the Orowan mechanism. Then, the corresponding hardness increase can be written as ∆𝐻𝑉𝑝𝑟𝑒𝑐𝑖𝑝 = 0.81𝐶𝑀𝐺𝑏𝑙𝑛(𝑑𝑚𝑒𝑎𝑛 /𝑏)/[2𝜋𝑑𝑚𝑒𝑎𝑛 (√1 − υ )(0.615√2𝜋/3𝑉𝑓 − 1)],
(4)
where C is a constant (C=3.06～3.28 for Al alloys[20,21], and Cmean=3.16 is adopted); M is the mean orientation factor (M=3.06 for the face centered cubic (f.c.c) Al matrix)[20,22];G is the shear modulus (G=27 GPa for 7000 series Al alloys)[23,24]; b is the Burgers vector (b= 0.286 nm for the f.c.c Al)[23,24]; υ is the Poisson’s ratio (υ=0.33 for the Al)[25]; Vf and dmean are respectively the volume fraction and the mean diameter of precipitates, which are listed in Table 3. The results were obtained using equation (3) and Table 3 and followed the order of ΔHVprecip(300°C) > ΔHVprecip(400°C) > ΔHVprecip(500°C) (ΔHVprecip(300°C)=23.041HV, ΔHVprecip(400°C)=22.682HV, ΔHVprecip(500°C)=21.234HV). Thus, the results of this approach correspond well with the measured hardness in a stable state (Fig. 8(a)), which further supports the present model. In a complex alloy system, multiple strengthening mechanisms are effective, including grain boundary strengthening, dislocation strengthening, precipitation strengthening and solid solution strengthening. A recent study revealed that for a 7075 Al alloy, grain boundary strengthening is the dominant strengthening mechanism[26]. For the 7N01 alloy, although the original material was rolled, the material did not recrystallize due to the short holding time during thermal cycling, taking into account that the grain size of the material during natural aging kept invariant. Thus, it is reasonable to state that the grain boundary strengthening is a relatively small contributor to the total strength variation in this study. Dislocation 8
strengthening is mainly affected by the processing deformation degree of the alloy. For the same processing state materials, the contribution of dislocation strengthening to the total strength could be considered as same. In the present study, it is comfirmed that both precipitation strengthening and solid solution strengthening are dominant mechanisms for 7N01 during natural aging after thermal cycle. According to previous research reported by KaKa[26], the solid solution strengthening is a relatively small contributor compare to the precipitation strengthening in 7000 series alloy. Moreover, the precipitation would reduce during the later natural aging. For the contribution of precipitation strengthening to the yield stress, following the Orowan equation [27]: 𝑝𝑟𝑒𝑐𝑖𝑝
∆𝜎0.2
= 0.4𝑀𝐺𝑏 ln(2𝑟𝑚𝑒𝑎𝑛 /𝑏) /(𝜋𝑠√1 − 𝑣 ).
(5) 𝑝𝑟𝑒𝑐𝑖𝑝
Likewise, the yield strength increments were arranged in the sequence of ∆𝜎0.2 𝑝𝑟𝑒𝑐𝑖𝑝
∆𝜎0.2
(300℃)＞
𝑝𝑟𝑒𝑐𝑖𝑝 (400℃)＞∆𝜎0.2 (500℃), which is consistent with the tendency of the curve in Fig. 8(b).
3.4 Microstructure of precipitation observed by TEM analysis
Fig. 9. TEM images showing distribution of strengthening precipitates at different magnifications: a) 500°C 9
and b) and c) 400°C As evidenced by the TEM images in Fig. 9, the strengthening mechanism for 7N01 is precipitation strengthening. Dislocation movement is controlled by the distribution of these precipitates, which act as pinning points (Fig. 9(a) and (b)). Precipitates exhibiting a symmetrical crescent shape are observed in Fig. 9(c) , which appear to be cut from the center line. According to previous research reported by Wang [28], this feature indicates that the dislocation encountered two barriers from shearable and non-shearable particles during movement. Finally, as the size, volume fraction, and spacing of precipitates have significant effects on the mechanical and electrical properties of a material, these properties will be discussed in the following paragraphs. Fig. 10 is the HRTEM images of the precipitates enclosed by the red circle in Fig. 9(c). The precipitate was identified as hexagonal η (MgZn2) phase with an orientation relationship to the Al matrix of (1120)η ∥(220)Al and (0002)η∥(002)Al. Fig. 10. (c) and (e) are the atomic model and simulated diffraction pattern of MgZn2 viewed from [1100]η, respectively, which has a hexagonal lattice with a=5.223Å and c=8.556Å (crystal data proposed by Wandahl et al. [29]). Obviously, It is in relatively good agreement between the observed structure (Fig. 10 (b) and (d)) and the atomic model (Fig. 10 (c) and (e)). Therefore, this precipitate structure can be identified as MgZn2. In addition, the Inverse fast Fourier transformation (IFFT) images in Fig. 10.(f) and (g), which were obtained from the (220) Al/ (2 2 0) Al and(002)Al/(002)Al diffraction patterns, respectively, clearly shows the existence of the dislocation along the interface between the η precipitate and the Al matrix. The lattice misfit between (1120)η and (220)Al is 45.3% ( d (1120) η = 0.2611nm, d (220) Al = 0.1427 nm, calculated as δ =(d (1120) η － d(220)Al)/d(1120)η). Hence, the interface between particle and matrix was non-coherent.
Fig. 10. (a) HRTEM images of the η(MgZn2) precipitates viewed from [110]Al; (b) the IFFT pattern after array mask of the framed area in the HREM image of (a); (c) atomic model of MgZn2 viewed from [1100]η;(d) and (e) the FFT pattern of the region discussed and simulated diffraction pattern of MgZn2 10
viewed from [1100]η;(f) and (g) the IFFT pattern obtained using the (220)Al/(2 2 0)Al and(002)Al/(002)Al diffraction patterns of the region discussed.
Fig. 11. Bright-field TEM micrographs for different peak temperatures after natural aging for 10 weeks: a) 300°C, b) 400°C, and c) 500°C The bright-field TEM micrographs in Fig. 11 show the microstructural evolution of the natural aged 7N01 alloys after the welding thermal cycles. The microstructure of the alloy for the peak temperature of 300°C consisted of coarse precipitates in which η′ precipitates were homogeneously distributed (Fig. 11(a)). Fig. 11(b) shows that the distribution of the precipitates for the alloy with the peak temperature of 400°C was similar to that for the peak temperature of 300°C. However, a significant decrease in the particle size was observed for the peak temperature of 500°C. This distribution could result from mostly dissolution of η′ and η above 400°C during the thermal cycle and dynamic precipitation process (the cooling stage of the thermal cycle and subsequent natural aging period), leading to the nucleation and growth of GP zones, which can be corroborated by the schematic in Fig. 7. In combination with the volume fraction, Eq. (4) was used to estimate the contribution of the 11
precipitates to the hardness, and then, ΔHVprecip was calculated. The volume fraction of the precipitates outlined by Pellissier and Purdy [30] can expressed as: 2 𝑉𝑓 = 1.4𝜋𝑁𝑑𝑚𝑒𝑎𝑛 /(6𝐴),
(6)
where N is the number of precipitates in the observation area and A is the image area. For the spacing between precipitates, it was hypothesized that the effect of precipitates on the mechanical and electrical properties of the Al alloy should scale as the spacing between precipitates. Then, according to the work of Dixit et al. [31], the corresponding spacing can be written as 𝑠 = 𝑟𝑚𝑒𝑎𝑛 (√2𝜋/3𝑉𝑓 − √8/3 ).
(7)
Table 3 Statistical results of intragranular precipitates for different peak temperatures of thermal cycle Peak temperature(℃) 300 400 500 6 2 Image area(*10 nm ) 7.040 19.526 11.807 Number of particles 54±9 128±15 158±30 Mean particle radius(nm) 60.705±3.21 69.885±2.32 25.345±6.13 Area fraction of particle (%) 8.647±1.13 7.593±0.13 2.952±0.43 Volume fraction of particle (%) 8.166±0.509 9.359±0.512 2.436±0.711 Spacing(nm) 209.257±20.626 217.147±16.243 192.868±12.473
Fig. 12. Particle size distribution (measured from TEM data)for different peak temperatures of thermal cycle: a) 300°C, b) 400°C, and c) 500°C 12
The precipitate size and distribution evolution for different processing temperatures were evaluated from bright-field TEM images. Part of the dislocations around the precipitates may affect the contrast and make TEM observations difficult, leading to large statistical results. To allow for a quantitative bulk measurement of the mean size and fraction of precipitates, it was necessary to count at least two bright-field TEM images containing less dislocations. In addition, because an increase in the sample size can assure the randomness of the measured sample and thus result in a reasonable overall distribution, the counting was performed using low-magnification TEM images. The effect of peak temperature on strengthening precipitates was determined by measuring the area fraction of particles, which refers to the percentage of the image area occupied by particles. The calculated area fractions of particles are listed in Table 3, revealing a gradual decrease during the welding thermal cycle with increasing peak temperature from 300°C to 500°C. It is evident from Fig. 12 that the particle size followed an approximately normal distribution. Moreover, the peak frequency of the particle radius did not vary greatly with temperature from 300°C to 400°C; however, the particle radius was small at 500°C (r300=50–70 nm, r400=50–70 nm, r500=10–20 nm). In addition, compared with the results for the peak temperatures of 300°C and 400°C, the mean value was significantly reduced for 500°C, as summarized in Table 3. Notably, a high frequency of particle sizes of r300=10–20 nm and r400＞100 nm are observed in Fig. 12(a) and (b), respectively. This finding indicates that the cluster and GP zone formation, using atoms provided by the dissolution of GP zones and η′ during the welding thermal cycle, was promoted during natural aging with the 300°C peak temperature and that coarsening of η occurred during the welding thermal cycle for the peak temperature of 400°C (coarsening of η consumes atoms used to form the GP zones during natural aging). This estimate is in good agreement with the schematic in Fig. 7. For the peak temperature of 500°C, the average size, area fraction, and peak frequency of precipitates were observed to decrease significantly. These results may be attributable to the (1) dissolution temperature of original large precipitates being reached during the welding thermal cycle, (2) destruction of precipitates by dislocation movement, and (3) formation of new small precipitates during the natural aging period. From the previous discussion (Fig. 7), the solvus temperature of η′ is below 300°C, thereby supporting option (1). The TEM observations in Fig. 9(c) revealed the symmetrical crescent shape of the particles, which is evidence of dislocation cutting. The dislocations passed through the second-phase particles and only slipped when the external stress exceeded the internal stress of the second-phase particles. The specimen in this experiment was only affected by the heat effect during processing and not by plastic deformation. Because the cutting was not caused by thermal cycling and the original material was subjected to deformation by rolling, option (2) can be eliminated. In addition, the peak frequency of the particle radius for the peak temperature of 500°C was r500=10–20 nm, which indicates that new precipitates originated from the GP zones during cooling of the thermal cycling and natural aging period, thereby validating the third option. Hence, it can be concluded that the severe deterioration of the properties of the evaluated particles with progressing peak temperature can be mainly attributed to precipitate dissolution. 4. Conclusion The microstructural evolution and mechanical behavior of a 7xxx alloy after welding thermal cycling and late natural aging were investigated. The effect of different peak temperatures of the welding thermal cycle on the precipitates and physical performance of Al alloys was analyzed, and the following conclusions were drawn: (1) High heating rates lead to the hysteresis effect of the reaction during DSC scanning; thus, the solvus temperature range of precipitates of 7N01 alloy during welding thermal cycling is higher than that measured using DSC. The solvus temperature of η′ was determined to be between 200°C and 300°C during the 13
welding process. Moreover, the formation and coarsening temperature of η was close to 400°C and the solvus temperature of η was between 400°C and 500°C during the welding process. (2) The Al–Zn–Mg–Cu alloy processed by thermal welding cycle exhibited obvious softening behavior; however, the hardness and strength were improved during natural aging. In addition, electrical conductivity measurements indicated that the SCC resistance was enhanced during the natural aging period. (3) Quantitative experimental data obtained from bright-field TEM images of the Al–Zn–Mg–Cu alloy revealed good consistency between the statistical indicators and physical properties during the natural aging period. Therefore, this work indicates that this reasonable statistic methodology could be introduced to predict the changes of the physical properties of these materials. Acknowledgements The present study was supported by the National Key Research and Development Program of China (Grant No. 2016YFB0700505). The authors thank Dr. D. C. Zhu and Mr. R. J. Huang from the Materials College of Sichuan University for their help with the electrical conductivity tests. References [1] Sachdev, A. K., Mishra, R. K., Mahato, A., & Alpas, A. (2012). Vehicle Lightweighting: Challenges and Opportunities with Aluminum. ICAA13: 13th International Conference on Aluminum Alloys. John Wiley & Sons, Inc. [2] Grong, Ø. (1997). Metallurgical modelling of welding. Institute of Materials. [3] B. I. Bjørneklett, Ø. Grong, O. R. Myhr, & A. O. Kluken. (1999). A process model for the heat-affected zone microstructure evolution in al-zn-mg weldments. Metallurgical & Materials Transactions A, 30(10), 2667-2677. [4] Sato, Y. S., Kokawa, H., Enomoto, M., & Jogan, S. (1999). Microstructural evolution of 6063 aluminum during friction-stir welding. Metallurgical & Materials Transactions A, 30(9), 2429-2437. [5] Hansen, V., Karlsen, O. B., Langsrud, Y., & Gjønnes, J. (2013). Precipitates, zones and transitions during aging of al-zn-mg-zr 7000 series alloy. Materials Science & Technology, 20(2), 185-193. [6] Zhang, S., Hu, W., Berghammer, R., & Gottstein, G. (2010). Microstructure evolution and deformation behavior of ultrafine-grained al–zn–mg alloys with fine η ′ precipitates. Acta Materialia, 58(20), 6695-6705. [7] Deschamps, A., Fribourg, G., Bréchet, Y., Chemin, J. L., & Hutchinson, C. R. (2012). In situ evaluation of dynamic precipitation during plastic straining of an al–zn–mg–cu alloy. Acta Materialia, 60(5), 1905-1916. [8] Deschamps, A., & Geuser, F. D. (2013). Quantitative characterization of precipitate microstructures in metallic alloys using small-angle scattering. Metallurgical & Materials Transactions A, 44(1), 77-86. [9] Dumont, M., Lefebvre, W., Doisneau-Cottignies, B., & Deschamps, A. (2005). Characterisation of the composition and volume fraction of η ′ and η, precipitates in an al–zn–mg alloy by a combination of atom probe, small-angle x-ray scattering and transmission electron microscopy. Acta Materialia, 53(10), 2881-2892. [10] Hutchinson, C. R., Geuser, F. D., Chen, Y., & Deschamps, A. (2014). Quantitative measurements of dynamic precipitation during fatigue of an al–zn–mg–(cu) alloy using small-angle x-ray scattering. Acta Materialia, 74(74), 96-109. [11] Liang, Z., Li, X., Nie, Z., Hui, H., Sun, J., & Sun, Z. (2016). Microstructure and mechanical properties of joints of a new al-zn-mg-cu alloy welded by tig. Rare Metal Materials & Engineering. 14
[12] Deschamps, A., Livet, F., & Bréchet, Y. (1998). Influence of predeformation on ageing in an al–zn–mg alloy—i. microstructure evolution and mechanical properties. Acta Materialia, 47(1), 281-292. [13] Buha, J., Lumley, R. N., & Crosky, A. G. (2008). Secondary ageing in an aluminium alloy 7050. Materials Science & Engineering A, 492(1–2), 1-10. [14] Jr, A. F. O., Barros, M. C. D., Cardoso, K. R., & Travessa, D. N. (2004). The effect of rra on the strength and scc resistance on aa7050 and aa7150 aluminium alloys. Materials Science & Engineering A, 379(1–2), 321-326. [15] Park, J. K., & Ardell, A. J. (1991). Microchemical analysis of precipitate free zones in 7075-a1 in the t6, t7 and rra tempers. Acta Metallurgica Et Materialia, 39(4), 591-598. [16] Li, X. M., & Starink, M. J. (2000). Analysis of Precipitation and Dissolution in Overaged 7xxx Aluminium Alloys using DSC. Materials Science Forum (Vol.331-337, pp.1071-1076). [17] Pouraliakbar, H., Jandaghi, M. R., & Khalaj, G. (2017). Constrained groove pressing and subsequent annealing of al-mn-si alloy: microstructure evolutions, crystallographic transformations, mechanical properties, electrical conductivity and corrosion resistance. Materials & Design, 124, 34-46. [18] Raeisinia, B., Poole, W. J., & Lloyd, D. J. (2006). Examination of precipitation in the aluminum alloy aa6111 using electrical resistivity measurements. Materials Science & Engineering A, 420(1–2), 245-249. [19] Chen, Y., Gao, N., Sha, G., Ringer, S. P., & Starink, M. J. (2016). Microstructural evolution, strengthening and thermal stability of an ultrafine-grained al–cu–mg alloy. Acta Materialia, 109, 202-212. [20] Xiao Guang Qiao, Nong Gao, & Marco J. Starink. (2012). A model of grain refinement and strengthening of al alloys due to cold severe plastic deformation. Philosophical Magazine, 92(4), 446-470. [21] Qiao, X. G., Starink, M. J., & Gao, N. (2009). Hardness inhomogeneity and local strengthening mechanisms of an al1050 aluminium alloy after one pass of equal channel angular pressing. Materials Science & Engineering A, 513(11), 52-58. [22] Taylor, G. I. (1934). The mechanism of plastic deformation of crystals. part i. Theoretical. Proceedings of the Royal Society of London, 145, 362-387. [23] Ma, K., Wen, H., Hu, T., Topping, T. D., Isheim, D., & Seidman, D. N., et al. (2014). Mechanical behavior and strengthening mechanisms in ultrafine grain precipitation-strengthened aluminum alloy. Acta Materialia, 62(5), 141-155. [24] Chen, Y., Gao, N., Sha, G., Ringer, S. P., & Starink, M. J. (2015). Strengthening of an al-cu-mg alloy processed by high-pressure torsion due to clusters, defects and defect-cluster complexes. Materials Science & Engineering A, 627, 10-20. [25] Wang, S. C., Lefebvre, F., Yan, J. L., Sinclair, I., & Starink, M. J. (2006). Vppa welds of al-2024 alloys: analysis and modelling of local microstructure and strength. Materials Science & Engineering A, 431(1–2), 123-136. [26] Ma, K., Hu, T., Yang, H., Topping, T., Yousefiani, A., & Lavernia, E. J., et al. (2016). Coupling of dislocations and precipitates: impact on the mechanical behavior of ultrafine grained al–zn–mg alloys. Acta Materialia, 103(15), 153-164. [27] Wen, H., Topping, T. D., Isheim, D., Seidman, D. N., & Lavernia, E. J. (2013). Strengthening mechanisms in a high-strength bulk nanostructured cu–zn–al alloy processed via cryomilling and spark plasma sintering ☆. Acta Materialia, 61(8), 2769-2782. [28] Wang, X. (2005). The shearable–non-shearable transition in al–mg–si–cu precipitation hardening alloys: implications on the distribution of slip, work hardening and fracture. Philosophical 15
Magazine, 85(26-27), 3113-3135. [29] Wandahl, G., Christensen, A. N., Uggerud, E., Songstad, J., Lönnberg, H., & Colacio, E., et al. (1989). Occurrence of extinction correlated with crystal growth mode for crystals of pd3ce and mgzn2. Acta Chemica Scandinavica, 43(3), 296-297. [30] Pellissier, G. E., Purdy, S. M. (1972). Stereology and Quantitative Metallography. American Society for Testing & Materials. [31] Dixit, M., Mishra, R. S., & Sankaran, K. K. (2008). Structure–property correlations in al 7050 and al 7055 high-strength aluminum alloys. Materials Science & Engineering A, 478(1–2), 163-172.
16