Improved creep aging response in Al-Cu alloy by applying pre-aging before pre-straining

Fangpu LYU,Chunhui LIU,*,Peipei MA,Jianshi YANG,Longhui CHEN,Lihua ZHAN, Minghui HUANG

a State Key Laboratory of Precision Manufacturing for Extreme Service Performance, School of Mechanical and Electrical Engineering, Central South University, Changsha 410083, China

b Light Alloy Research Institute, Central South University, Changsha 410083, China

c Advanced Research Center, Central South University, Changsha 410083, China

KEYWORDS Al-Cu alloy;Creep aging;Pre-aging;Precipitation;Pre-strain

Abstract The improvement of post-form properties without compromising creep formability has been a critical issue in creep age forming of aluminum alloy component.A pretreatment process incorporating artificial pre-aging at 165 °C for 6 h/12 h/24 h followed by pre-strain (3%–9%)has been developed.This method not only evidently improves the strength but also accelerates the creep deformation during creep aging of an Al-Cu alloy.A strength increase of 50 MPa with a slight decrease of ductility relative to the 9%pre-strained alloy is acquired in the alloy artificially pre-aged for 24 h regardless of the pre-strain level (3%–9%).Artificial aging for 24 h prior to 3%pre-strain enables an increase of creep strain by 30%.The creep strain in the alloy artificially preaged for 24 h and pre-strained by 6% is comparable to that in the alloy pre-strained by 9%.The strength and ductility in the alloy artificially pre-aged for 6 h/12 h and pre-strained by 3%are even slightly higher than those in the alloy purely pre-strained by 9%.The characterizations by transmission electron microscopy reveal that pre-aging at 165 °C could promote the accumulation of dislocations during pre-straining due to the pinning effect of pre-existing Guinier-Preston zones (GP zones)/θ′′ phases and thus expedite the creep deformation in respect to the pure pre-straining treatment.The enhanced precipitation of θ′ phases at these pinned dislocations contributes to the improved strength after creep aging.The results demonstrate applying artificial pre-aging before pre-straining is an efficient strategy to elevate the creep aging response in Al alloys.

1.Introduction

Creep Age Forming(CAF),originally developed for manufacturing extra-large structural parts of aluminum alloys in the aeronautic industry,1has attracted increasing interests as a promising technology to obtain high mechanical properties and low residual stress.2More details for the development of the CAF can be found in a review.2The creep aging process depends on the stress-state3,4and initial tempers3,4and determines the springback of the Al alloy component.5CAF allows for simultaneous forming and strengthening of externally loaded sheet metal at elevated temperatures through creep/stress relaxation and age hardening, respectively.6The elastic deformation is transformed into irreversible deformation by a creep/stress relaxation mechanism, while the strength of the alloy is increased mainly by solid-state precipitation and marginally by dislocation proliferation during CAF.7,8AA2219 aluminum alloys are widely used in aerospace applications for their high specific strength, good machinability and good corrosion resistance to make aircraft wing bulkheads and rocket fuel tanks.9–11However,the relatively low strength compared to 7xxx series and Al-Cu-Li based 2xxx series alloys limits their applications.12,13Therefore, the simultaneous improvement of creep formability and post-form strength is of critical importance for the development of CAF to produce high-performance AA2219 aluminum alloy components.

Various pre-treatment methods have been developed to enhance the creep aging response in Al alloys.Yang et al.14,15experimentally investigated the effects of pre-deformation on creep strain, mechanical properties and microstructure of the as-quenched AA2219 alloy.The results show that predeformation can prolong the duration of the initial creep stage and significantly increase the creep strain.The mechanical properties are also significantly improved due to the heterogeneous precipitation of θ′phase at the pre-deformation induced dislocations.16–18Liu et al.9,19proposed a strategy to adjust the density and mobility of dislocations generated by cold rolling during creep aging to improve the creep formability and post-form mechanical properties of aluminum alloys at low cost.Tuning dislocations is thus promising to obtain better creep formability and performance during CAF by fundamentally altering the creep aging behavior in AA2219 alloy.Lu et al.20investigated the changes in the dislocation organizations and properties of AA2219 aluminum alloy during Artificial Aging (AA) with the degree of pre-deformation.They found that pre-deformation increased the dislocation density in the alloy matrix and thus accelerated the precipitation of dense, fine and uniformly distributed nano-phases, resulting in an improved strength of the alloy.Liu et al.21found that applying cold rolling after AA could lead to a higher strengthening potential in the Al alloys than those after sole waterquenching or Natural Aging(NA).This is ascribed to the formation of finer and more homogeneous precipitates compared to those pretreated by NA.Chen et al.22discovered that creep formability and the overall properties of the formed alloy were improved by a combination of pre-aging and rolling predeformation compared to the conventionally treated Al-Mg-Si alloy.Therefore, the introduction of a pre-aging process prior to pre-deformation is promising to further improve the creep formability and the mechanical properties of Al-Cu alloys.However, the effect of combined pre-aging and prestrain on the creep aging response in Al-Cu alloys remains unexplored.

In the present work, the influence of artificial pre-aging at 165 °C for 6–24 h plus pre-straining by 3%–9% on the creep aging behavior and mechanical properties of an Al-Cu alloy has been systematically investigated.The through-process evolution of dislocations and precipitation phases in the alloy is studied in detail using transmission electron microscopy.This work aims to develop an effective processing route for enhancing the creep aging response in Al-Cu alloys, as could also be transferrable to other Al alloy series.

2.Materials and experimental procedures

The raw material used was commercial AA2219 aluminum alloy sheets with a thickness of 10 mm, and its chemical composition is shown in Table 1.Dog-bone shaped specimens with a specification length of 35 mm and a diameter of 5 mm were machined along the rolling direction of the received plate.

The material preparation and testing procedures are schematically shown in Fig.1.The alloy was first solid solution treated at 535 °C for 45 min to obtain a supersaturated solid solution,8,23and then water quenched to room temperature.The alloys were subjected to Natural Pre-Aging (NPA) and Artificial Pre-Aging (APA) within 5 min after water quenching, respectively.The APA was conducted based on the conventional T6 tempers11,16,24and various durations were used to tune the size/density of the pre-existing precipitates.APA at 165°C for 6–24 h and NPA for 2 days to reach the T4 state were adopted to obtain two different kinds of pre-aging treatments.The pre-aging was followed by a pre-strain stretching by 3%, 6%, and 9% on a SUST-5105 universal tensile testing machine at an initial strain rate of 2 mm/min.The pretreatments are summarized in Table 2.After the pretreatments were completed, the specimens were subjected to uniaxial tensile creep tests on an RMT-D10 creep tester with an accuracy of±2°C and±3 N.After heating the specimens to the target temperature of 165 °C, a stress of 150 MPa was applied to the specimens and held at constant stress for 12 h.This representative creep aging condition was selected according to the literature and is recognized to enable the optimum balance between peak-aged strength and creep strain.8,19Quasi-static uniaxial tensile tests were performed at room temperature with an initial strain rate of 8.33×10–4/s on a 100 kN MTS Landmark machine.Repeated tests for each processing condition are performed at least three times to ensure the accuracy of the data.

The Transmission Electron Microscopy (TEM) samples were first mechanically ground to 80 μm, and then punched into disks with a diameter of 3 mm.The disks were finally electropolished in an electrolytic solution of 30vol% nitric acid and 70vol% methanol9,25under a voltage of 15 V, with the cooling temperature kept below–25°C.Microstructure observations were performed using a Talos F200X transmission electron microscope operated at 200 kV and the precipitates were detected with the e-beam along the 〈001〉A(chǔ)l direction under the High-Angle Annular Dark-Field Scanning Transmission Electron Microscopy (HAADF-STEM) mode.

Table 1 Chemical composition of AA2219 aluminum alloy.

Fig.1 Schematic diagram of experimental procedures for 2219 alloy with different pre-treatments.

Table 2 Summary of pre-treatment processes.

3.Results and analyses

3.1.Creep deformation under different pre-aging and pre-strain conditions

The creep deformation behavior of AA2219 alloy under different pre-treatment conditions is shown in Fig.2 for comparison.The creep strains for different pre-treatments are also presented in Table 3.The creep strain curves were first smoothed to eliminate experimental errors, and then differentiated with time to calculate the strain rate curves.Figs.2(a)–(c) show the creep strain curves of Al-Cu alloy with different pre-aging conditions under the same pre-strains, while Figs.2(d)–(f) show the corresponding creep rate curves.All curves are characterized by two creep stages including initial creep and steady-state creep.During the first 4 h (Figs.2(a)–(c)),the creep strain increases rapidly with a high creep strain rate.This is caused by the dislocations generated under the applied stresses at the elevated temperature.2,26The dislocations can move in the alloy due to the low barrier resistance at this stage and thus lead to a rapid increase in creep deformation.7,15However, the gradual formation of precipitates and the accumulation of dislocations reduce the mobility of the dislocations, giving rise to a drastic decrease in the creep strain rate.As the creep aging proceeds, a steady creep stage is reached (4–12 h) due to the dynamic balance of recovery of dislocations and strain hardening.2,8,27At this time, the creep rate is reduced to a very low value and keeps constant, while the creep strain increases slowly and linearly.

Both the creep rate and the creep strain are larger in the APA alloy with a 3% pre-strain than its NPA counterpart,demonstrating the beneficial effect of the APA pre-treatment.After APA treatment for 6 h, 12 h and 24 h, the final creep strain increases from 0.125% to 0.175%, 0.168% and 0.164%, respectively.For the alloy pre-strained by 6%, the total creep strains of the APA alloy are almost the same as the NPA alloy.However, the creep strain rate improves pronouncedly with the increasing APA time.Such a high creep strain rate at the beginning of creep aging is advantageous for shortening the CAF processing time.When the pre-strain level is increased to 9%,the creep strain of the APA alloy after 12 h decreases slightly compared to the NPA alloy and the reduction enlarges with the increasing APA time.Nevertheless,the creep strain and the creep strain rate at the primary creep stage in the APA alloy, APA 6 h-9%sample in particular, are obviously bigger than those in the NPA alloy.At creep aging temperatures, the dislocations introduced by pre-strain can accelerate the dynamic precipitation and growth of nanophases in the alloy.8,28As creep aging proceeds, the growth and coarsening of the precipitates can affect the dislocation movement, reducing dislocation mobility and resulting in a lower creep rate of the alloy.Such a phenomenon should be even more pronounced in the APA alloys.This may explain the slight decrease of the creep strain in the later stage of creep aging when APA is applied before 6%–9% pre-strained alloy.Nevertheless, the APA alloy is able to possess a high creep strain during the first 4–8 h.As shown in Figs.2(b), (c), and Table 3, the creep strain of the APA 24 h-6% alloy equals or even surpasses that of the NPA-9% alloy in the first 4 h.This indicates that APA treatment can improve the creep formability with reduced pre-strain, as may benefit the pre-straining process in the practical manufacturing.In general, the APA could enhance the pre-straining effect on the creep deformation in terms of the total creep strain at small pre-strain level(3%) and initial creep strain at higher pre-stain level (6%–9%).Note that both the creep strain and the strength in the 2219 alloy without pre-strain are much lower than those in the pre-strained 2219 alloy.14,15The creep strain of NPA alloy without pre-strain after 12 h at 165 °C under 150 MPa is less than 0.1%.9As the precipitates formed by APA could impede the dislocation motion,the creep strain in the APA alloy without pre-strain is expected to be even lower.

3.2.Tensile test results after creep aging

The tensile properties of the differently pre-treated Al-Cu alloy samples after Creep Aging (CA) are presented in Fig.3.The results of representative samples are shown to emphasize the pre-aging effect.For the conventional T4(NPA)alloy,increasing the pre-strain level could improve the yield strength and tensile strength of the creep-aged alloy with a ductility loss.For the alloy with an APA time of 24 h,the mechanical properties are similar in terms of both strength and ductility for the three samples with different pre-strains (3%, 6% and 9%).Compared with the NPA treatment, the yield strength of the creep-aged APA alloy with the same pre-strain ratio is increased by at least 50 MPa and a uniform elongation above 7% is preserved.A yield strength increase of 100 MPa is acquired in the creep-aged APA 24 h-3% alloy relative to the NPA-3% alloy.After creep aging, it is notable that only 3% pre-strain after APA of 6 h could impart a strength and a uniform ductility both slightly higher than those in the conventional NPA-9% alloy.Prolonging the APA time further improves the creep-aged strength in the 3% pre-strained alloy and even a yield strength of 400 MPa is obtained in the APA 24 h-3% alloy.This suggests that the APA treatment prior to pre-strain can effectively improve the mechanical properties of the Al-Cu alloy after creep aging.

Fig.2 Creep deformation at 165 °C under 150 MPa for 12 h in samples with different pre-treatments.

3.3.Microstructure characterization

In order to understand the microscopic mechanism underlying the effect of APA pretreatment on the creep aging of the investigated alloy,the samples including NPA-6%,NPA-9%,APA 24 h-6% and APA 24 h-9% were selected for microstructure characterization.As shown in Fig.4 (the photographs were acquired with the electron beam along[001]Al),pre-strain with higher level introduces more dislocations inside the Al-Cu alloy before creep aging is performed.Under the same prestrain level, the dislocation density in the Al-Cu alloy treatedby the pre-aging process is larger.The dislocations tend to pile up in the pre-aged alloy,as is also in stark contrast to the uniform distribution of dislocations in its counterpart without pre-aging.This suggests that pre-aging promotes the accumulation of dislocations during pre-strain.The dislocation densities in different samples are also roughly estimated based on the Taylor equation29σρ=MαGbρ1/2,where σρis the strength increment caused by pre-strain, M is the Taylor factor for the Al matrix (usually taken as 3.06), α is a constant that is approximately 0.3, G is the shear modulus (about 27 GPa), b is the magnitude of the Burgers vector(about 0.286 nm for aluminum),ρ is the estimated dislocation density.The values of ρ are shown in Table 4.

Table 3 Summary of creep strains of 2219 alloy with different pretreatments.

The results are in good accordance with the TEM observations that APA induces more dislocations during pre-straining than its NPA counterpart.During creep aging, the density of movable dislocations is believed to determine the creep rate.9Therefore,this may explain the higher creep rates in the initial stage of creep aging in the pre-aged and pre-strained alloy, as will be discussed in detail later.

Fig.3 Mechanical properties of creep-aged Al-Cu alloy with different pre-treatments.

Fig.4 TEM photographs showing dislocation distribution in pre-treated alloy before creep aging.

Table 4 Summary of dislocation density of 2219 alloy with different pretreatments.

For the pre-strained age-hardening aluminum alloys, the precipitation phases evolve accompanying the dislocations and play an important role in the macroscopic properties.16,18,30The generally accepted precipitation sequence in Al-Cu alloys is Supersaturated Solid Solution(SSSS) → Guinier-Preston zones (GP zones) → θ′′→ θ′(Al2Cu),16where θ′′is the metastable intermediate phase that is completely coherent with the matrix, and θ′is the relatively stable and semi-coherent precipitate with thin plate-like morphology.All these phases lie on three equivalent {001}Al planes.These metastable phases impede the motion of dislocations during plastic deformation, with the θ′phase being considered as the main strengthening phase of Al-Cu alloys.17,31The study of the effect of the pretreatment on the precipitation phase evolution is thus essential.

Fig.5 shows the TEM photographs of the four selected samples after creep aging.The photographs were acquired with the electron beam along the [001]Al.The insets are Selected Area Diffraction Patterns (SADPs) revealing clear diffraction spots (indicated by arrows) from precipitates.It can also be noticed that the number of θ′precipitates increases with the increase of pre-strain level.The dislocations in the APA alloy mostly remain after creep aging.By comparing the NPA alloy with the APA alloy, it is evident that the latter ones have a higher density of dislocations in the matrix after creep aging.After creep aging for 12 h,a mixture of large θ′phase and fine θ′′phase is formed in the matrix of the pre-strained alloy.A notable difference is that the NPA alloy has a much larger percentage of θ′′precipitates after creep aging, while the APA alloy mainly contains θ′precipitates.This is also reflected by the diffraction spots in the inset of Fig.5.In Figs.5(c) and(d), the dislocations are pinned by the dense θ′precipitates,restricting the movement of dislocations.This phenomenon becomes more obvious with the increase of pre-strain.This may be the reason for the decrease of creep rate at late stage of creep aging and also for the slight loss of ductility in the APA alloy after creep aging, although the dislocation density in the NPA alloy appears to be lower after creep.In the naturally pre-aged alloys,the dislocations accumulated during prestraining should be more likely to recover during creep aging.Considering that the four samples underwent different pretreatments (i.e.NPA and APA), the nucleation and growth of θ′precipitates as well as the influence of dislocations during creep aging should be varied.

Fig.5 TEM photographs showing morphology and dislocation distribution in creep-aged Al-Cu alloy.

In order to investigate the effect of different pre-treatments on the evolution of precipitates during creep aging, HAADFSTEM characterizations were carried out before and after creep aging of the NPA-9% and APA 24 h-6% samples, as shown in Fig.6, the insets are high-resolution photographs of the main precipitate in the corresponding sample.The photographs were acquired with the electron beam along the[001]Al.Apart from the dislocation segments introduced by 9% pre-strain, it can be found that a large number of extremely fine GP I zones of only a few nanometers in size exist in the NPA alloy due to natural aging.The artificial preaging induces the formation of numerous θ′′phases, which could be sheared during subsequent pre-straining (Fig.6(b)).There is no precipitation of the shear-resistant θ′phase before creep aging.This is reasonable because the artificial aging temperature of 165°C is not sufficient to activate the nucleation of the θ′precipitates in the pre-strain-free alloy.32,33As presented in Figs.6(c) and (d), the shearable phase θ′′and the shearresistant phase θ′coexist in the alloy creep aged at 165 °C under 150 MPa.The pre-existing dislocations have been reported to promote the nucleation and growth of θ′-type precipitates,16,18,34,35leading to a bimodal precipitation of hardening particles containing fine θ′′precipitates in between large θ′precipitates.This kind of microstructure contributes to the high strength while ensuring good plasticity of the Al-Cu alloy.8Contrary to the apparent bimodal precipitation in the NPA alloy, the θ′precipitates dominate in the creep-aged APA alloy, as is manifested by the much higher percentage of θ′precipitates in Fig.6(d).More dislocations are accumulated in the alloy that underwent the artificial pre-treatment and thus enhance the heterogeneous nucleation of θ′.On the other hand, the modification of the θ′′phase imposed by prestraining should also have accelerated the transition to θ′.The higher percentage of θ′could pin dislocations (Figs.5(c)and (d)) and thus may explain the low creep rate in the APA alloy at the late creep aging stage.

Fig.6 HAADF-STEM photographs showing morphology and distribution of precipitates before and after creep aging of Al-Cu alloy samples.

Fig.7 Size distribution of θ′ precipitates after creep aging of differently pre-treated alloys.

As shown in Fig.7,the size and percentage of precipitation phases after creep aging are quantitatively analyzed for the APA 24 h-6%and NPA-9%samples.At least 200 precipitates were measured in each sample to acquire the required statistical accuracy.The types of precipitated phases are the same for the two different samples (Fig.6).The size distribution of the precipitated phases of the main strengthening phase θ′in the alloy is different to each other.For the creep-aged NPA-9%sample, the θ′precipitates have a broader distribution and a larger average diameter (about 60 nm) than those in the APA 24 h-6% sample (with an average diameter of about 45 nm).This is due to the uniform precipitation of fine-sized θ′after creep aging of the APA 24 h-6% alloy (Fig.6(d)).The relative frequency of θ′and θ′′is shown in Fig.7(c).The APA 24 h-6% alloy has a higher frequency of θ′precipitates after creep aging.This demonstrates that APA 24 h before pre-strain could promote the precipitation of θ′during creep aging, as may increase the percentage of θ′and decrease the average size of θ′.The pinning effect of large number of θ′precipitates increases the strength of the alloy while decreasing the creep rate through reducing the dislocation mobility at the later stage of creep aging.

The creep strain at the initial stage of the APA 6 h-9%alloy is the highest among the investigated samples, and the creepaged strength is also obviously higher than the conventionally treated samples (Figs.2 and 3).Therefore, detailed characterization of the APA 6 h-9%-CA sample was carried out, as shown in Fig.8.The inset of Fig.8(a) is a SADP revealing clear diffraction spots(indicated by arrows) from precipitates.Similarly, after creep aging, the alloy still has many dislocations.Compared to the APA 24 h-9% sample, the mobility of the dislocations should be larger due to the lower impeding effect of the pre-existing precipitates.Therefore, the higher density of mobile dislocations gives rise to the higher creep rate in the APA 6 h-9%sample.Both the SADP and the HAADFSTEM photographs (Figs.8(b) and (c), where the inset are high-resolution pictures of the main precipitate in the corresponding sample.The photographs were acquired with the electron beam along the [001]Al) show that a large amount of θ′precipitates (about 40 nm in diameter) exist in this alloy.The θ′phases are uniformly distributed in the alloy matrix and θ′′phases have a low number density.This demonstrates that artificial pre-aging for 6 h could already have a considerable promoting effect on the precipitation of θ′phase during creep aging.

4.Discussion

4.1.Enhanced creep aging response by artificial pre-aging before pre-strain

Improving the creep aging response in the Al-Cu alloy is critical for creep age forming of components with high precision and high performance.The results demonstrate artificial preaging results in an increase of both creep strain and strength at the small pre-strain level (3%).In particular, the combination of APA 6–24 h and 3% pre-strain could obtain a strength-ductility balance better than that in the 9% prestrained NPA sample.A yield strength of 400–420 MPa could be acquired in the pre-strained samples pre-aged for 24 h.By comparison, the T8 tempered 2219 alloy has a yield strength of only 360–380 MPa.9,36,37Such a high strength is advantageous for reducing the weight of the Al alloy structural parts used in the aerospace industry.

Minitab software has been used to analyze the range and variance of the measured experimental data.The sensitivity of two factors, i.e.pre-aging condition and pre-strain, on the yield strength,tensile strength and elongation of the alloy after creep aging has been investigated.The experimental results and data analysis are shown in Tables 5–7,respectively,where R is a value representing the range while F (the mean square value of each factor divided by the mean square of the error)represents the variance.Larger values of R and F represent higher influence of the corresponding factors.The results of range and variance are consistent.It is found that the preaging condition has the greatest effect on the mechanical properties of the alloy, while the pre-strain treatment possesses the second highest effect.

According to the classical Orowan equation,38,39the strain rate is determined by the dislocation movement as ˙ε=ρmb v-,where ˙ε is the creep strain rate, ρmis the density of movable dislocations, v- is the average rate of dislocation motion.Both ρmand v- evolve upon creep aging, leading to the change of creep strain rate.For the pre-strained NPA and APA alloys,the difference in v- is marginal at initial stage of creep aging.Thus the initial creep rate is mainly dictated by the magnitude of ρm.This is consistent with the TEM observations that the APA alloy stores more dislocations after pre-strain compared to the NPA alloy.Increasing the pre-strain level to 6%–9%could induce a higher dislocation density and thus further enhance the initial creep strain rate.However, with the proceeding creep aging, more shear-resistant θ′precipitates that effectively impede dislocation motion are formed in the alloy matrix (Figs.6–8).This situation is more severe in the APA alloy pre-strained by 6%–9% and results in a significant decrease of the creep strain rate.Consequently,the APA alloy pre-stained by 6%–9%usually has a higher creep strain during the first 4 h of creep aging(Fig.2(c))while a lower creep strain afterwards.Increasing the pre-strain level to 6%–9% could add to the density of movable dislocations in the alloy matrix and thus enhance the creep strain during creep aging.It is notable that APA prior to pre-strain could further improve the creep strain,especially at the early stage of the creep aging.Moreover,the APA 24 h sample with a pre-strain of 6%could achieve a creep strain comparable to the NPA-9% sample before the first 4 h of creep aging (Fig.2(c)).Therefore, artificial pre-aging followed by pre-strain proves to be an effective method to raise the creep aging response.

Fig.8 TEM photographs showing distribution of dislocations and precipitates in APA 6 h-9%-CA alloy.

Table 5 Input experimental results for regression analysis of factor sensitivity.

Table 6 Results of range analysis of experimental data.

Table 7 Results of analysis of variance for experimental data.

The steady state creep is determined by a dynamic balance between the hardening due to dislocation generation and the dislocation recovery.27However, the situation is much more complex in the creep aging of pre-strained Al-Cu alloy due to the coupling of multiple factors.Firstly,the pre-existing dislocations affect the dislocation evolution together with the newly generated dislocations during creep.Secondly, the continuous solutes precipitation and its interaction with the dislocations (e.g.dislocation enhanced precipitation of θ′phase)have a fundamental effect on the dislocation movement.Note that the primary creep strain constitutes a large fraction of the total creep strain in pre-strained alloy (Fig.2).During this stage,the resistance to dislocation motion by precipitates is relatively small and the recovery should play a critical role.There is complex interaction between creep deformation and precipitation hardening during creep aging.Progressive growth of hardening precipitates like θ′phase could hinder the dislocation movement carrying the creep deformation.Our findings reveal there is an evident decline of creep rate in the APA alloy relative to the NPA alloy at the late stage of creep aging,though the APA alloy has a higher creep rate at early stage.This indicates that the artificial pre-aging before pre-strain also profoundly modifies the precipitation process during creep aging.In addition, though artificial pre-aging could increase the accumulated dislocations in the pre-strained alloy, further artificial pre-aging also reduces the dislocation mobility.As shown in Fig.5, despite the relatively higher dislocation density in the APA alloy, most of the dislocations are pinned by the growing θ′,reducing the motility of the dislocations.Therefore, the creep rate of APA alloys has a substantially lower vand a relatively lower creep stain during the late creep aging.

4.2.Pre-treatment dependent microstructural evolution during creep aging

The analysis in Section 4.1 reveals that the APA could improve the creep deformation and mechanical properties of the Al-Cu alloy compared to the conventional NPA treatment.To understand the underlying mechanism, the through-process microstructures are observed and analyzed in detail.Based on the experimental results, the microstructure evolution is summarized and illustrated in Fig.9.The Al-Cu alloy was first solid solution treated and quenched to generate a supersaturated solid solution.The alloy was then subjected to natural pre-aging and artificial pre-aging, respectively.The natural pre-aging occurs at room temperature and is inevitable during the manufacturing process of the Al alloys.HAADF-STEM observations reveal that single Cu-atom layer GP I zones with a diameter of 3–5 nm uniformly distribute in the Al matrix.The artificial pre-aging introduces a large number of platelike θ′′phases with a diameter of about 20 nm.These θ′′phases are generally composed of more than two Cu atom-layers separated by three Al atom-layers.During subsequent pre-strain,dislocations with different densities and distributions are introduced.More dislocations accumulate in the APA alloy and a proportion of θ′′precipitates appear to be cut through by dislocations.This suggests that the APA alloy has a larger strain hardening ability than the NPA alloy during pre-straining due to the difference in the pre-existing precipitates.Movable dislocations benefit the creep deformation in the Al alloy.15The increased dislocation density results in higher creep rate in the early stage of creep aging for the APA alloy than that for the NPA alloy with the same pre-strain level.For each type of sample(i.e.NPA and APA samples),the dislocation density and thus the creep strain increase with the pre-strain level(Fig.2).

Fig.9 Schematic diagram of microstructure evolution during creep aging of pre-aged and pre-strained Al-Cu alloy.

The difference in pre-existing dislocations and precipitates fundamentally alters the microstructure evolution during creep aging.As creep aging proceeds, the nucleation and growth of the θ′precipitates would occur heterogeneously at the stored dislocations in the matrix,20leading to a gradual strength increase in the alloy.During creep aging, the θ′precipitation phase grows faster and can be clearly visualized in both the naturally pre-aged and artificially pre-aged alloy.In addition,the GP zones and the shearable θ′′phases precipitate homogeneously in the matrix.The θ′precipitates in the APA alloy account for a much higher percentage compared to the NPA alloy, indicating the favorable growth of θ′phase in the APA alloy.The reasons for the faster precipitation kinetics in the APA alloys are two-fold.Firstly, the pre-strained APA alloy contains more dislocations that can accelerate the atom movement according to the short-circuit diffusion.Secondly,a large number of meta-stable θ′′phases in the matrix are cut through by dislocations and could transform in-situ to more stable θ′phases.40,41This is reasonable considering the high-density θ′phases in the creep-aged APA alloy.The transition from θ′′to θ′phase is normally difficult to occur.Nevertheless, the stress field at the pinned dislocation around θ′′precipitate could promote such a transition.42The excessive vacancies in the pre-strained alloy are also believed to facilitate the formation of θ′phase.40Therefore, the percentage of θ′is relatively higher for APA alloys than that for NPA alloys after creep aging.

The creep-aged APA alloy could achieve a pronounced strength increase in respect to its NPA counterpart.This is attributed to the dominance of θ′phase which is more effective in strengthening than the θ′′phase,e.g.the APA 6 h-3%alloy can even exceed the strength of the NPA-9%alloy(Fig.3 and Table 5).Slightly larger ductility can also be obtained in the APA alloy with lower pre-strain because much less dislocations pre-exist,as is another advantage.However,the preferential growth of θ′precipitates would reduce the dislocation mobility and thus the creep rate, as could explain the lower creep strain at the late stage in the APA alloy (Fig.2).The bimodal distribution of the precipitates and the retained dislocations together improve the strength of the alloy, while the presence of the shearable phase benefits the plasticity of the alloy.21

5.Conclusions

Creep tests,tensile tests and TEM characterizations have been performed to study the effect of combined artificial pre-aging and pre-straining on the creep aging properties of Al-Cu alloy.The combined pre-treatment includes artificial pre-aging at 165 °C for 6 h/12 h/24 h followed by pre-strain (3%–9%).The main conclusions are made as follows:

(1)The initial creep rate of the pre-treated alloy is increased regardless of the pre-aging time.The creep strain of the APA 24 h-6% alloy equals or even surpasses that of the NPA-9%alloy in the first 4 h.Relative to the NPA alloy,the APA alloy increased the total creep strain by 30% at low pre-strain level(3%)while slightly decreased the total creep strain at high prestrain level (9%).Compared to the NPA sample, the strength of creep-aged APA sample with the same pre-strain level increases by 50–100 MPa while maintains a good ductility.The strengths of all the APA samples pre-strained by 3%alloy are higher than that of NPA sample pre-strained by 9%.

(2) The θ′′phases generated by APA promote the accumulation of dislocations due to the pinning effect, leading to the increase of the initial creep rate.A bimodal precipitation of θ′and θ′′phases is observed in both the creep-aged NPA and APA samples.The enhancement of the precipitation of θ′phase by artificial pre-aging increases the mechanical strength while decreases the creep rate at the late stage of creep aging.

(3) The varying creep deformation and mechanical properties in pre-aged and pre-strained Al-Cu alloy demonstrate that pre-existing precipitates and dislocations fundamentally alter the microstructure evolution during creep aging.Applying artificial pre-aging before pre-strain has the potential to synergistically improve the creep formability and post-form properties for CAF of Al-Cu alloys.The newly developed method may also be applicable to the other aluminum alloy series.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

The authors thank the financial support from the National Natural Science Foundation of China (Nos.52274404,U2032117, U22A20190), Natural Science Foundation of Hunan Province, China (No.2022JJ20065), the Science and Technology Innovation Program of Hunan Province, China(No.2022RC1001)and the National Key Research and Development Program of China (No.2021YFB3400903).