Recrystallization Behavior of 7175 Al Alloy during Modified Strain Induced Melt Activated (SIMA) Process

Recrystallization Behavior of 7175 Al Alloy during Modified

Strain-Induced Melt-Activated (SIMA) Process

Young Buem Song

_{, Kyung-Tae Park}

_{and Chun Pyo Hong}

1

Department of Metallurgical Engineering, Yonsei University, Seoul 120-749, Korea

2_{Division of Advanced Materials Science and Engineering, Hanbat National University, Taejon 305-719, Korea}

A modified strain-induced, melt-activated (SIMA) process for semi-solid processing of alloys was proposed. In order to examine the applicability and effectiveness of the modified SIMA process, the recrystallized microstructures of a high strength 7175 Al alloy prepared by the modified SIMA processes were macroscopically compared to that of conventional process. The modified SIMA process employed casting, two stage homogenization, warm multi-forging, and recrystallization and partial melting (RAP) instead of the conventional process consisted of casting, hot working, cold working and RAP. For the alloy processed by the conventional SIMA process, the recystallized grain size decreased with increasing the cold rolling reduction ratio up to 20% and then almost unchanged. RAP treatments of the 20% cold rolled alloy at above 600_{C and for longer than 30 min resulted in significant grain growth. In case of the modified SIMA process, the alloy multi-forged with the}

accumulated strain of about 7 and RAP at 575_{C for 10 min exhibited the uniform equiaxed recrystallized grain structure similar to that of the}

conventional SIMA process under the optimum conditions. The improved processing efficiency of the modified SIMA process over the conventional one,i.e.removal of hot working and saving of RAP time, is attributed to the enhanced recrystallization kinetics due to a high density of Mg(Zn,Al,Cu)2precipitates which are formed by two stage homogenization, acting as the preferential recrystallization sites and due

to the driving force imposed by relatively large amount of uniform deformation by multi-forging.

(Received January 20, 2006; Accepted March 10, 2006; Published April 15, 2006)

Keywords: semi-solid processing, strain induced melt activated (SIMA) process, homogenization, multi-forging, recrystallization, aluminum alloy

1. Introduction

Semi-solid processing (SSP)1)in which metal forming is

carried out in a mushy zone provides very effective assess for

near net shaping of metals and alloys.2)The success of SSP

strongly depends on the rheological characteristics of materials affected by their microstructures, such as phase fraction, grain size and shape, phase network state, etc., at

given processing conditions.3,4) _{In general, it is known that}

the skeleton structure of fine solid spheroids without

dendrites is the most suitable for SSP.5)_{In order to produce}

such sound initial microstructure for SSP, several techniques such as magneto-hydrodynamic (MHD) casting and spray

casting have been developed and commercialized.6–8)_Since

liquid is cooled to a mushy zone in MHD casting and spray casting, these processes are difficult to apply to the alloys which are sensitive to hot cracking and inadequate to economic fabrication of the small-size semi-solid billets. In order to overcome such deficiencies, an alternative process called the stress induced, melt activated (SIMA) process was

developed.9)

The SIMA process involves the four processing stages: casting, hot working, cold working, and recystallization and

partial melting (RAP).9,10) _{In the SIMA process, an alloy}

having the sufficient stored energy by cold working is heated to a mushy zone. During heating, recovery and recrystalliza-tion occur before liquid formarecrystalliza-tion with the aid of the stored energy. Reaching the mushy zone, liquid is formed by preferential melting at grain boundaries with high energy state, and penetrates into high angle boundaries of recrystal-lized grains. Accordingly, the amount and distribution of the stored energy by cold working is the most critical factor in the SIMA process since they control the recovery and

recrystallization kinetics and the uniformity of the resultant microstructure.

The present investigation was motivated by two main considerations. First, the distribution of the stored energy by cold rolling is usually inhomogeneous across the work piece

section,11) _{causing non-uniform microstructure at the RAP}

stage. Therefore, the alternative working techniques which can distribute the large amount of the stored energy more homogeneously throughout the work pieces should be employed to ensure the microstructure uniformity. Second, in case of the difficult-to-work alloys, the amount of cold rolling is limited in order to avoid cracking; for instance, about 20% for the high strength wrought Al alloys. In such a case, the selection of optimum RAP conditions becomes important to obtain the desired skeleton structure with fine equiaxed grains for successful SSP. Accordingly, it is essential to figure out the effects of RAP conditions on the microstructures of the alloys having the cold working limit in the conventional SIMA process.

The purpose of this study is two-fold; (a) introduction of a modified SIMA process and its application to a high strength 7175 Al alloy, and (b) examination of the separate effect of each stage of the conventional SIMA process on recrystal-lization of the same alloy in a mushy zone. A modified SIMA process includes homogenization and warm multi-forging instead of hot working and cold working in the conventional SIMA process (Fig. 1). Multi-forging is an effective process

for microstructure refinement.12–16) _{In order to examine the}

effectiveness of the present new process, the RAP micro-structures developed by both conventional and modified processes were compared.

2. Experimental Procedures

For the present investigation, a high strength 7175 Al alloy

*_{Graduate Assistant}

(Al-5.6Zn-2.5Mg-1.6Cu-0.23Cr (in mass%)) with 380 mm was direct-chill-cast. The liquidus, solidus and eutectic

temperatures of the alloy were 635, 532 and 477C,

respectively.

For the conventional SIMA process, the direct-chill-cast

alloy was homogenized at 460_{C for 30 h, and then hot rolled}

to the thickness of 12 mm at 420_{C. The hot rolled plates}

were further subjected to cold rolling and subsequent RAP followed by water quenching. In order to examine the effect of each stage on recrystallization, the reduction ratio of cold rolling and RAP temperature and time were varied

inde-pendently: (a) 5–32% reduction followed by RAP at 575_C

for 30 min, (b) RAP temperatures of 400–625_{C for 30 min}

with 20% reduction ratio, and (c) RAP times of 1–1440 min

at 575_{C with 20% reduction ratio.}

For the modified SIMA process, the round bar samples of

60 mm90 mm taken from the homogenized billets were

rehomogenized at 482_{C for 3 h and then furnace cooled. The}

precipitates are anticipated to form and coarsen during the two stage homogenization and to act as the effective sites for large stored energy accumulation during subsequent

multi-forging. After reheating the rehomogenized billets to 250_C,

warm multi-forging was conducted by a 200 ton hydraulic press equipped with a flat die. Multi-forging consisted of repetitive upsetting and down-drawing. Each cycle of upsetting and down-drawing was designed to yield an

accumulated true strain (") of 0.8–0.9: "is the sum of

ln (Ao=A) of each upsetting and down-drawing whereAoand

Aare the initial and final cross-sectional area of the sample,

respectively. The pictorial illustration of the present

multi-forging and the photos of the sample multi-forged to "¼

7:0and 9.5 are shown in Fig. 2. RAP in the modified SIMA

process was performed at 575 and 600_{C for 10 min.}

The microstructures were observed by optical microscopy (OM) and scanning electron microscopy (SEM, Philips XL30-FEG). The recrystallized grain size was measured by an image analyzing software (WINATceh, BMI plus SE) and the precipitates were identified by X-ray diffractometer (XRD, Jeol JDX 8030) and energy disperse spectroscopy (EDS).

3. Results and Discussion

3.1 Conventional SIMA process

3.1.1 Effect of cold working

In order to examine the effect of cold working amount on RAP microstructure, the hot rolled plates were cold rolled

with the reduction ratios of 5–32% and then recrystallized at

575_{C for 30 min. The representative OM micrographs of}

RAP microstructure are shown in Fig. 3. Without cold rolling (0%), RAP microstructure consisted of equiaxed grains with the presence of liquid phase at some grain boundaries. However, the grain size distribution was bimodal, small

grains of 20mm and large grains of 100mm. The grain

size was steadily decreased up to 15% cold rolling. By further cold rolling, the grain size remained nearly unchanged in association with dynamic recovery, and its distribution became more uniform.

The effect of cold rolling on the RAP microstructure can be explained in terms of the grain

deformation-recrystalliza-tion model.17)_{As the cold rolled alloy is still in the solid state}

during heating to the mushy zone, recovery and recrystalli-zation occur at the expense of the stored energy. By holding at the mushy zone, liquid formation occurs preferentially at high angle grain boundaries (HAGBs) of recrystallized grains since these HAGBs are in the higher energy state than the strain free recrystallized grain interiors. Concurrently, liquid penetrates into HAGBs by a capillary effect, causing the fragmentation and/or enclosure of grains. The increment of cold rolling amount accelerates recovery and tion kinetics and provides the larger number of recrystalliza-tion sites. Therefore, the more severe cold rolling is imposed, the finer grain size is obtained.

3.1.2 Effect of RAP temperature

The 20% cold rolled alloy was recrystallized in the

temperature range of 400–625_{C for 30 min and the resultant}

microstructures are shown in Fig. 4. As discussed in the preceding section, the RAP experiments were conducted on the 20% cold rolled samples since cold rolling over 20% may be harmful due to cracking for the present alloy without further grain refinement. The cold rolled microstructure

remained up to 500_{C,i.e.no recrystallization. The partially}

recrystallized microstructure appeared at 525C. Fully

recrystallized equiaxed grains of 50–60mm were obtained

at 550 and 575_{C. At 600C, significant grain growth}

T emper ature T emper ature (a) (b) T T RT S L T T RT S L Hot rolling Cold rolling WQ WQ Recrystallization & Partial melting Recrystallization & Partial melting FC Re-homogenization multiple forging (warm working)

Fig. 1 Schematic illustration of (a) conventional SIMA process, and (b) new modified SIMA process.

(a)

(b)

shaping for upsetting upsetting down-drawing

Σε = 9.5 Σε = 7.0

occurred and most grains were surrounded by liquid layer.

The microstructure at 625C consisted of coarse grains with

localized liquid pools.

The semi-solid with 50–70% solid can bear its own weight and so exhibits the solid-like behavior. When it is subject to

shear deformation, it behaves as high viscosity liquid.18)In

100 µm

10 %

20 % 30 %

0 %

0 10 20 30

0 20 40 60 80 100

recr

ystalliz

ed g

ra

in siz

e

,

d

/

µ

m

cold rolling reduction ratio (%)

Fig. 3 Variation of the recrystallized grain size with the cold rolling reduction ratio (RAP at 575_{C for 30 min) in conventional SIMA}

processed 7175 Al alloy.

100 µm

500 C 525 C

550 C 575 C

600 C 625 C

Fig. 5, the solid fraction of the alloy is predicted as a function

of the RAP temperature by the Scheil’s equation19)_of

fS¼1

TMT

TMTL

1=ð1kÞ

ð1aÞ

Vf ¼ fs

fSþ ð1fSÞðS=LÞ

ð1bÞ

where fSandVf are the weight and volume fraction of solid

respectively, TM, TL, and T are the melting, liquidus, and

RAP temperatures, respectively,SandL are the solid and

liquid density, respectively, and k is the partitioning

coefficient. As shown in Fig. 5, the solid fraction at 575

and 600_{C are80 and 60%, respectively. Therefore, RAP at}

600_{C seems to be desirable for SSP but it should be}

reminded that a coarse grain structure at 600_{C as revealed in}

Fig. 4 deteriorates the rheological properties by lowering the flow resistance of semi-solid during shear deformation of SSP.20)

3.1.3 Effect of RAP time

The microstructural change of the 20% cold rolled alloy

with RAP time at 575_{C is shown in Fig. 6.}

Recrystallized grains appeared at 5 min. Beyond 10 min, microstructure was mostly covered with recrystallized equiaxed grains and grain growth occurred continuously with increasing time by the Ostwald ripening mechanism. During Ostwald ripening, small grains with large boundary curvature remelt into liquid and large grains grow by receiving solute atoms from liquid with higher solute

content.21–26)_{When grain growth is controlled by the volume}

diffusion of solute atoms in the solid state, the grain growth behavior generally obeys the third power non-ideal grain

growth law27,28)of

d3d_o3¼Kt ð2Þ

wheredis the grain size at timet,dois the initial grain size,

andK is the grain growth rate coefficient. In order to check

the applicability of eq. (2) to the present case, d3

(¼d3_d

o3) is plotted against RAP time in Fig. 7. As seen

in Fig. 7, the data exhibit the excellent linearity, indicating that grain growth of the alloy in the semi-solid state was rate-controlled by volume diffusion.

3.2 Modified SIMA process

3.2.1 Microstructure developed by rehomogenization

and multi-forging

A representative SEM micrograph of the alloy after

rehomogenization at 482_{C for 3 h followed by furnace}

cooling is shown in Fig. 8. After rehomogenization, two types of phases were observed; one coarse boundary phases (‘A’ in Fig. 8) and the other rod-shaped precipitates at grain interior (the area ‘B’ in Fig. 8). By EDS and XRD analysis,

the former was identified as eutectic Al7Cu2Fe or Al2CuMg

intermetallic compounds with low melting point.29)_{These are}

probably formed at interdendrites by solute rejection during

partial melting. The latter was identified as Mg(Zn,Al,Cu)2

precipitated during rehomogenization. These rod-shaped

Mg(Zn,Al,Cu)2 precipitates with the length of 5mm are

enough to act as the preferential sites for stored energy accumulation during subsequent multi-forging, and, in turn,

for recrystallization during RAP.30) _{The microstructures}

developed by warm multi-forging at 250_{C are shown in}

525 550 575 600 625 650

0.0 0.2 0.4 0.6 0.8 1.0

f_s = 0.7

f_s = 0.5

f_s = 0.59 at 600°C f_s = 0.79 at 575°C

F

raction of Solid

Temperature, T/°C 7175 Al

T_S=536°C

T_L=635°C

Fig. 5 Variation of the solid fraction with RAP temperature for a 7175 Al alloy predicted by the Scheil’s Eq.14)

100 µm

5 min 30 min

180 min 60 min

Fig. 6 Optical micrographs of the RAP microstructures at various RAP times (20% cold rolling and RAP at 575_{C) in conventional SIMA}

Fig. 9. The number and size of coarse boundary phases (the black patches in Fig. 9) were considerably reduced and their distribution became homogeneous by repetitive multi-forg-ing, resulting in relatively uniform microstructure. In

addition, the coarsening of Mg(Zn,Al,Cu)2 precipitates

concurrently occurred due to warm multi-forging

temper-ature of 250C.

3.2.2 RAP microstructure

After multi-forging with "¼7:0 and 9.5, RAP was

conducted at 575 and 600_{C for only 10 min and the resultant}

RAP microstructures are shown in Fig. 10. Without

multi-forging, the grain size was 130mm regardless of RAP

temperature. In addition, the shape of most grains was out of spheroid and the liquid pools locally existed at some grain boundaries. This liquid localization was mainly attributed to the first melting of coarse boundary eutectic phases with low melting point since no strain-induced melting was expected. By imposing multi-forging, the grain size was considerably

refined. The grain size at 575C was slightly smaller than that

at 600_{C. It was quite noticeable that the grain size developed}

by only 10 min RAP of the modified SIMA process (50mm)

was comparable to that produced by 30 min RAP of the

conventional SIMA process (55mm). The RAP time saving

of the modified SIMA process resulted from the enhanced recovery and recrystallization kinetics due to imposition of relatively large deformation by multi-forging and uniform

distribution of a large number of recrystallization sites,i.e.

Mg(Zn,Al,Cu)2precipitates, formed by rehomogenization.

3.2.3 Advantages of the modified SIMA process

The present modified SIMA process exhibits several advantages over the conventional one. First, the hot working stage in the conventional SIMA process is removed and so the modified process can be applied to the alloys sensitive to hot working. Second, more uniform microstructure can be

A

A A B

area ‘B’

Fig. 8 A representative SEM micrograph of the alloy and the EDS profile of the rod shaped precipitates after rehomogenization at 482_{C for 3 h}

followed by furnace cooling in new modified SIMA processed 7175 Al alloy.

0 200 400 600 800 1000 1200 1400 1600

0 1x106 2x106

3x106

4x106 5x106 6x106 7x106

∆

d

3 /µ

m

3

Time, t/min

K=4605 R2

=0.9959

Fig. 7 A plot ofd3_{vs RAP time showing the validity of volume diffusion}

controlled grain growth.

40 µm (a)

(b)

(c)

achieved by homogeneous distribution of fine precipitates and uniform deformation by multi-forging. Third, the RAP time is remarkably saved by enhanced recrystallization kinetics associated with a high density of fine precipitates acting as the preferential recrystallization sites and relatively large amount of deformation by multi-forging. Finally, the semi-solid billets can be fabricated without the dimensional limit, while the conventional SIMA process employing cold rolling cannot produce the heavy sectioned plates or bars due to cold cracking. Even if extrusion is employed instead of rolling, cold extrusion is extremely difficult especially for large size billets.11)

4. Conclusions

(1) In the present investigation, the separate effects of each stage (cold working, and RAP temperature and time) of the strain-induced, melt-activated (SIMA) process on recrystallization of a high strength 7175 Al alloy for its semi-solid processing were investigated. In addition, to overcome the deficiencies of the conventional SIMA process, a modified SIMA process in which homoge-nization and warm multi-forging were employed instead of hot working and cold working in the conventional one was proposed.

(2) For the alloy processed by the conventional SIMA process, the recystallized grain size decreased with increasing the cold rolling reduction ratio up to 20% and then remained nearly unchanged with further cold

rolling. RAP treatments of the 20% cold rolled alloy

above 600_{C and longer than 30 min resulted in}

significant grain growth. The grain growth behavior of the alloy well obeyed the third power grain growth law, indicating the volume diffusion controlled growth. (3) The alloy which was prepared by the modified SIMA

process of multi-forging with the accumulated strain of

7 and RAP at 575_{C for 10 min exhibited the uniform}

equiaxed grain structure similar to that of the alloy processed by conventional SIMA process with 20%

cold rolling and RAP at 575_{C for 30 min.}

(4) The improved processing efficiency of the modified

SIMA process over the conventional one,i.e. removal

of hot working and RAP time saving, results from the accelerated recrystallization kinetics associated with a

high density of Mg(Zn,Al,Cu)2precipitates formed by

rehomogenization acting as the preferential recrystalli-zation sites and with the driving force for recrystalliza-tion by a large amount of uniform deformarecrystalliza-tion by multi-forging.

REFERENCES

1) D. B. Spencer, R. Mehrabian and M. C. Flemings: Metall. Trans.3

(1972) 1925–1932.

2) Z. Franz: Int. Mater. Rev.47(2002) 49–85. 3) M. C. Flemings: Metall. Trans.22A(1991) 957–981.

4) E. Tzimas and A. Zavaliangos: Acta Mater.47(1999) 517–528. 5) E. Tzimas and A. Zavaliangos: Mater. Sci. Eng.A289(2000) 228–240. 6) Metals Handbook, vol. 15, 9th _{ed. ASM, Metals Park (OH, USA)}

200 µm 575 C

575 C

575 C 600 C

600 C

600 C

Σε Σε = 0

Σε Σε = 0

Σε

Σε = 7.0 Σε Σε = 7.0

Σε Σε = 9.5

Σε Σε = 9.5

(1988) 327–338.

7) P. S. Grant: Prog. Mater. Sci.39(1995) 497–545.

8) W. R. Loue and M. Suery: Mater. Sci. Eng. A203(1995) 1–13. 9) K. P. Young, C. P. Kyonka and J. A. Courtois: US Patent 4, 415, 374

(1983).

10) D. H. Kirkwood, C. M. Sellars and L. G. Elias Boyed: European Patent 0305375 B1 (1992).

11) H. V. Atkinson: Prog. Mater. Sci.50(2005) 341–412.

12) A. Belyakov, T. Sakai, H. Miura and R. Kaibyshev: Scripta Mater.42

(2000) 319–325.

13) O. Sitdikov, T. Sakai, A. Goloborodoko, H. Miura and R. Kaibyshev: Phil. Mag.85(2005) 1159–1175.

14) R. Z. Valiev, R. K. Islamgaliev and I. V. Alexandrov: Pro. in Mater. Sci.45(2000) 103–189.

15) V. M. Imayev, R. M. Imayev and G. A. Salishchev: J. Mater. Sci. Eng. A208(1996) 226–231.

16) M. Noda, M. Hirohashi, K. Funami and Y. Suwahara: J. Japan Inst. Metals66(2002) 101–108.

17) R. D. Doherty, H. Lee and E. Feest: Mater. Sci. Eng. A65(1984) 181–

189.

18) D. H. Kirkwood: Int. Mater. Rev.39(1994) 173–189.

19) B. K. Dhindaw, L. Kumar, N. C. Amer Alkarkhi and H. Fredriksson: Mater. Sci. Eng. A413–414(2005) 156–164.

20) E. Tzimas and A. Zavaliangos: Mater. Sci. Eng. A289(2000) 217–227. 21) J. M. Lifshitz and V. V. Slyozov: J. Phy. Chem. Solids19(1961) 35–

50.

22) C. Wagner: Z. Electrochem.65(1961) 581–591.

23) A. M. Kliauga and M. Ferrante: Acta Mater.53(2005) 345–356. 24) G. W. Greenwood: Acta Metall.4(1956) 243–248.

25) L. Anestiev and L. Froyen: Phy. Rev. E61(2000) 6721–6731. 26) S. C. Hardy and P. W. Voorhees: Met. Trans. A19(l988) 2713–2721. 27) E. D. Manson-Whitton, I. C. Stone, J. R. Jones, P. S. Grant and

B. Cantor: Acta Mater.50(2002) 2517–2535.

28) S. Annavarapu and R. D. Doherty: Acta Metall.43(1995) 3207–3230. 29) C. Huang and S. Kou: Welding J. Oct. (2002) 211–222.