A
Applied
Y. Birol: Effect of solution heat treatment on the age hardening capacity of dendritic and globular AlSi7Mg0.6 alloys
Yucel Birol
W 2010 Carl Hanser Verlag, Munich, Germany
www.ijmr.de
Not for use in internet or intranet sites. Not for electronic distribution.
Materials Institute, Marmara Research Center, TUBITAK Gebze, Kocaeli, Turkey
Effect of solution heat treatment on the age hardening capacity of dendritic and globular AlSi7Mg0.6 alloys The effect of solution heat treatment on the age hardening capacity of dendritic and globular AlSi7Mg0.6 alloys was investigated in the present work to find out if solution treatment times are in fact shorter for rheocast parts. The morphology of the a-Al matrix, whether dendritic or globular, has no effect on the response of the alloy to T6 treatment. A solution treatment at 540 8C for 2 h suffices to fully solutionize the Mg and Si and the age hardening capacity of the AlSi7Mg0.6 alloy is impaired when it is 1 h or shorter. This is true for both dendritic and globular alloys. The expected favourable impact of the globular structure on the solution treatment times is believed to be offset by the relatively coarser structure in the globular alloy with respect to its dendritic counterpart. Keywords: Semi-solid forming; Rheocasting; Aluminium alloys; Heat treatment
1. Introduction Semi-solid forming (SSF) is capable of producing components that can meet the stringent requirements of the automotive industry, by combining the near-net-shape capabilities of die casting with the mechanical properties of forging [1–4]. Parts formed in the semi-solid state are reported to be substantially higher in quality than die castings and lower in cost than forgings [5]. SSF is now a commercial manufacturing route which annually produces millions of near net-shape parts for the automotive industry [6]. SSF often relies on AlSi7Mg casting alloys [7], since globular grains suspended in a liquid matrix, a prerequisite for SSF [8], is readily obtained in these alloys owing to a relatively high volume of Al–Si eutectic. Besides their excellent castability, AlSi7Mg alloys offer good mechanical properties and corrosion resistance and have thus been the material of choice for a variety of automotive components [9–11]. The semi-solid processed AlSi7Mg components are generally heat treated to improve strength and ductility [12], often with those practices that are employed for conventionally cast alloys [13]. The former, however, may require heat treatment cycles markedly different from those employed for their dendritic counterparts, thanks to a unique uniform globular microstructure which means a relatively shorter diffusion path [13]. Int. J. Mat. Res. (formerly Z. Metallkd.) 101 (2010) 3
A number of researchers have investigated the heat treatment of aluminium components formed in the semi-solid state [13–17]. The majority of these investigations claimed that the semi-solid cast components enjoy improved mechanical properties in T6 and T5 tempers and/or shorter solution treatment cycles than their dendritic counterparts. The effect of solution heat treatment on the age hardening capacity of dendritic and globular AlSi7Mg0.6 alloys was investigated in the present work to find out if the solution treatment times are in fact shorter and if the age hardening capacity is higher in the latter.
2. Experimental procedures AlSi7Mg0.6 (357) alloy ingot with the chemical composition given in Table 1, was melted in an induction furnace set at 750 8C. The melt thus obtained was processed conventionally as well as via the semi-solid route: directly cast into a permanent mould to obtain dendritic samples and cooled into the semi-solid temperature range with the internal cooling process [18, 19] before casting to obtain globular structures. Samples sectioned from these ingots were soaked at 540 8C for 1 to 16 h and then quenched in hot water. These samples were subsequently submitted to artificial ageing treatment at 190 8C. As-cast and solution-treated samples were prepared with standard metallographic techniques: ground with SiC paper, polished with 3 micron diamond paste and finished with collodial silica. They were examined after etching with a 0.5 % HF solution using a Olympus BX51M model optical microscope. Hardness of the aged samples was measured in Vickers units and reported as the average of minimum 5 readings. Differential scanning calorimetry (DSC) was employed to identify the precipitation reactions. DSC runs were carried out by placing a 3.5 mm sample disc in the sample pan and a super purity aluminium of equal mass in the reference pan of the cell. The cell was equilibrated at 25 8C and then heated to 600 8C at 10 K min–1 in a dynamic Table 1. Chemical composition of the AlSi7Mg0.6 alloy used in the present investigation (wt.%). Si
Fe
Cu
Mn
Mg
Sr
Ti
6.87
0.181
0.0049
< 0.01
0.577
0.0114
0.141
439
A
Applied Y. Birol: Effect of solution heat treatment on the age hardening capacity of dendritic and globular AlSi7Mg0.6 alloys
Not for use in internet or intranet sites. Not for electronic distribution.
argon atmosphere (1 l h–1). The heat effects associated with transformation reactions were then obtained by subtracting a super purity Al baseline run from a given heat flow curve. The DSC curves obtained from the experiments were highly reproducible.
3. Results and discussion Microstructures of the dendritic and globular AlSi7Mg0.6 alloys in the as-cast state are illustrated in Fig. 1. a-Al dendrites and the interdendritic network of the Al–Si eu-
W 2010 Carl Hanser Verlag, Munich, Germany
www.ijmr.de
Fig. 1. Microstructure of (a, b) dendritic and (c, d) globular AlSi7Mg0.6 alloys in the ascast state.
Fig. 2. XRD patterns of the dendritic AlSi7Mg0.6 alloy (a) in the as-cast state, (b) after soaking at 540 8C for 8 h and (c) and finally annealing at 425 8C for 1 h. &: a-Al, .: Si, !: p-AlSiMgFe, ¤: b-Mg2Si.
440
Int. J. Mat. Res. (formerly Z. Metallkd.) 101 (2010) 3
A
Applied
Y. Birol: Effect of solution heat treatment on the age hardening capacity of dendritic and globular AlSi7Mg0.6 alloys
(Al8Si6Mg3Fe) and AlFeSi particles between a-Al dendrites and globules, respectively. The XRD spectrum exhibits, in addition to the reflections of Al and Si, relatively weaker p-AlSiMgFe lines (Fig. 2a). Few unindexed lines are believed to be generated by the AlFeSi intermetallics the volume fraction of which is apparently too small to al-
Not for use in internet or intranet sites. Not for electronic distribution.
tectic phase, typical of conventionally cast hypoeutectic Al–Si alloys, are readily identified in the dendritic alloy (Fig. 1a and b). Dendrites are replaced by relatively coarser a-Al globules in the rheocast counterpart (Fig. 1c and d). Both dendritic and globular alloys are dominated by eutectic Al–Si phase and additionally contain p-AlSiMgFe
W 2010 Carl Hanser Verlag, Munich, Germany
www.ijmr.de
Fig. 3. Microstructure of (a, c, e, g, i) dendritic and (b, d, f, h, j) globular AlSi7Mg0.6 alloys after solution treatment at 540 8C for (a, b) 1, (c, d) 2, (e, f) 4, (g, h) 8 and (i, j) 16 h.
Int. J. Mat. Res. (formerly Z. Metallkd.) 101 (2010) 3
441
A
Y. Birol: Effect of solution heat treatment on the age hardening capacity of dendritic and globular AlSi7Mg0.6 alloys
low identification with XRD. The morphological features of the AlFeSi particles suggest that they are predominantly of the monoclinic b-Al9Fe2Si2 variety. Several changes were noted upon solution treatment at 540 8C (Fig. 3). The Si particles of the eutectic phase have spheroidized and coarsened with increasing soaking time. This process involved shrinking and eventual disappearance of some Si particles at interdendritic eutectic regions and subsequent precipitation of the solute Si thus made available at the coarser particles. These activities led to the removal of some dendrite boundaries marked by the Si particle stringers and to the enrichment of the matrix with solute Si after 16 h at 540 8C as evidenced by the intragranular contrast that develops upon etching (Fig. 3i and j). With dendritic features gradually fading away, the structure of the dendritic alloy became more like that of the globular counterpart. Finally after 16 h at 540 8C, the features of the two alloys were quite similar (Fig. 3i and j). The eutectic regions were found to shrink also in the globular alloy while Si particles have spheroidized and coarsened with increasing soaking time. The minor phases have attained more compact morphologies in both dendritic and globular alloys. These structural changes were not at all evident from the XRD patterns of the solutionized samples which are almost identical to those of the as-cast counterparts (Fig. 2b). The response to artificial ageing of dendritic and globular alloys which have been submitted to solution treatment at 540 8C is illustrated in Fig. 4. The age hardening curves of the two morphologies are very similar regardless of the
soaking time, yielding nearly identical peak hardness and time-to-peak hardness values. The morphology of the a-Al matrix, whether dendritic or globular, apparently has no effect on the response of the alloy to T6 treatment. The soaking time, on the other hand, appears to have an impact on the age hardening capacity, in nearly the same fashion in both alloys. The maximum hardness values that can be achieved upon artificial ageing increases with increasing soaking time. The improvement is negligible, however, for soaking times above 2 h, all providing a minimum peak hardness in the neighbourhood of 120 HV within 2 h at 190 8C during subsequent artificial ageing. The peak hardness drops below 115 HV when soaking lasts only one hour. It is thus fair to conclude that soaking at 540 8C for 2 h suffices to fully solutionize the Mg and Si and that the age hardening capacity of the AlSi7Mg0.6 alloy is impaired when it is employed for 1 h or less. The age hardening response outlined above can be accounted for fairly accurately by the DSC analysis of samples submitted to solution treatment. With one endothermic and five exothermic signals, the DSC spectrums of the dendritic and globular alloys are nearly identical (Fig. 5). This peak configuration is typical of excess-Si wrought AlMgSi (6XXX) alloys [20–27] with the exception of signal 5, which is evidently linked with the much higher Si content of the casting alloys [28, 29].
W 2010 Carl Hanser Verlag, Munich, Germany
www.ijmr.de
Not for use in internet or intranet sites. Not for electronic distribution.
Applied
(a) (a)
(b) (b) Fig. 4. Hardening curves of (a) dendritic and (b) globular AlSi7Mg0.6 alloys artificially aged at 190 8C after soaking at 540 8C for different times.
442
Fig. 5. DSC scans of (a) dendritic and (b) globular AlSi7Mg0.6 alloys after soaking at 540 8C for different times. 1: formation of clusters; 2: dissolution of GP-I zones; 3: precipitation of b’’; 4: formation of b’; 5: precipitation of Si; 6: formation of b.
Int. J. Mat. Res. (formerly Z. Metallkd.) 101 (2010) 3
A
Applied
W 2010 Carl Hanser Verlag, Munich, Germany
www.ijmr.de
Not for use in internet or intranet sites. Not for electronic distribution.
Y. Birol: Effect of solution heat treatment on the age hardening capacity of dendritic and globular AlSi7Mg0.6 alloys
It is inferred from the precipitation sequence of AlMgSi alloys [20–27] that the formation of clusters and the dissolution of GP-I zones which have evolved from these clusters are responsible for the low-temperature exothermic and subsequent endothermic signals (signals 1 and 2), respectively. Signals 1 and 2 are increasingly larger with increasing soaking time. More of the available Mg and Si are solutionised in the matrix for clustering activities when the soaking step is longer. This is most evident, however, when the soaking time is increased from 1 h to 2 h with a much bigger signal 3 in the latter. Signal 3 is readily linked with the precipitation of the principal hardening phase, b’’. The b’’ peak is slightly displaced to higher temperatures in samples soaked at 540 8C for 2 h and more as it takes longer to dissolve a higher population of GP-I zones, and is bigger owing to a higher solute content made available for precipitation. The next three peaks seem to largely overlap with one another, thus making a triplet. The first of these peaks, signal 4, is linked with the transformation of the coherent b’’ to the semi-coherent b’ phase while the next exothermic peak (signal 5) was shown by the XRD analysis of samples heat treated at 350 8C, to be associated with Si precipitation [28]. Finally, signal 6 is produced by the transformation of the semi-coherent b’ phase to the equilibrium b variety. This is confirmed by the XRD analysis of samples annealed at 425 8C which shows, in addition to Al, Si and p-AlSiMgFe reflections, equilibrium b lines (Fig. 2c). The increase in the intensity of Si lines in Fig. 2c is due to Si precipitation between approximately 300 8C and 350 8C. The precipitation sequence of dendritic and globular AlSi7Mg0.6 alloys are almost identical and can be written as, SSS ? clusters ? GP-1 zones ? b’’? b’ + Si ? b, offering a significant age hardening capacity for both. The sharp increase in hardness during the first two hours of the ageing treatment is produced by the precipitation of the principal hardening phase, b’’. The start of the hardness drop after 4 h of ageing at 190 8C marks the onset of the transformation of the coherent b’’ phase to the semi-coherent b’ phase transformation followed by the precipitation of Si and the transformation of b’ to the incoherent equilibrium b-Mg2Si phase. The marked change in size of the b’’ peak when soaking time is increased from 1 to 2 h is consistent with the improvement in peak hardness of samples soaked for at least 2 h. It takes at least 2 h at 540 8C to fully solutionize the Mg and Si to offer the maximum hardening capacity during a subsequent artificial ageing cycle for both dendritic and globular AlSi7Mg0.6 alloys. Longer soaking times hardly provide any improvement in the peak hardness or time-topeak hardness values. The optimum solution treatment time at 540 8C for the dendritic alloy in the present work is then shorter than that reported for a A356 alloy in a recent investigation [30]. That for the globular alloy, on the other hand, is longer than the optimum solution treatment times at 540 8C reported for rheocast A356 alloy [14, 31]. It is not at all evident from the present work that shorter solution heat treatments can be employed for the globular AlSi7Mg0.6 alloy. Any favourable effect of a uniform equiaxed structure in the former is believed to be offset by its relatively coarser globule size. Caution must be exercised and the average globule size must be taken into consideration when solution treating globular alloys. ConsiderInt. J. Mat. Res. (formerly Z. Metallkd.) 101 (2010) 3
ing that the average globule size is almost always bigger in rheo-cast alloys (often > 50 microns) than the dendrite arm spacing in pressure die cast grades which enjoy much higher solidification rates, a shorter solution treatment in the former is claimed to be highly unlikely.
4. Conclusions The morphology of the -Al matrix, whether dendritic or globular, has no effect on the response of the AlSi7Mg0.6 alloy to T6 treatment. A solution treatment at 540 8C for 2 h suffices to fully solutionize the Mg and Si, providing a peak hardness in the neighbourhood of 120 HV within 2 h at 190 8C during subsequent artificial ageing. The age hardening capacity is impaired when soaking lasts 1 h or less. This is true for both dendritic and globular alloys. The expected favorable impact of the globular structure on the solution treatment times is believed to be offset by the relatively coarser structure in the globular alloy with respect to its dendritic counterpart. It is a pleasure to thank Mr. O. Çakır and Mr. F. Alageyik for their help with the experiments.
References [1] M.C. Flemings: Metall. Trans. A 22 (1991) 957. [2] D.H. Kirkwood: Int. Mater. Rev. 39 (1994) 173. [3] Z. Fan: Int. Mater. Rev. 47 (2002) 1. DOI:10.1179/095066001225001076 [4] M.P. Kenney, J.A. Courtois, R.D. Evans, G.M. Farrior, C.P. Kyonka, A.A. Koch, K.P. Young, in: Metals Handbook, Ninth Edition, 15 (1988) 327. [5] H. Wang, C.J. Davidson, D.H. StJohn: Mater. Sci. Eng. A 368 (2004) 159. DOI:10.1016/j.msea.2003.10.305 [6] M. Garat, S. Blais, C. Pluchon, W.R. Loue, in: A.K. Bhasin, J.J. Moore, K.P. Young, S. Midson (Eds.), Proceedings of the Fifth International Conference on Semi-Solid Processing of Alloys and Composites. Golden, Colorado (1998) 17. [7] G. Hirt, R. Cremer, A. Winkelmann, T. Witulski, M. Zillgen: J. Mater. Process. Tech. 45 (1994) 359. DOI:10.1016/0924-0136(94)90366-2 [8] W.R. Loue, M. Suery: Mater. Sci. Eng. A 203 (1995) 1. DOI:10.1016/0921-5093(95)09861-5 [9] G.S. Cole, A.M. Sherman: Mater. Charact. 35 (1995) 3. DOI:10.1016/1044-5803(95)00063-1 [10] D. Brungs: Mater. Design 18 (1997) 285. [11] P. Cavaliere, E. Cerri, P. Leo: J. Mater. Sci. 39 (2004) 1653. DOI:10.1023/B:JMSC.0000016165.99666.dd [12] M. Rosso, I. Peter, R. Villa: Sol. St. Phen. Vols. 141 – 143 (2008) 237. DOI:10.4028/www.scientific.net/SSP.141-143.237 [13] H. Möller, G. Govender, W.E. Stumpf: Sol. St. Phen. Vols. 141 – 143 (2008) 737. DOI:10.4028/www.scientific.net/SSP.141-143.737 [14] M. Rosso, M. Actis Grande: Sol. St. Phen. Vols. 116 – 117 (2006) 505. DOI:10.4028/www.scientific.net/SSP.116-117.505 [15] E. Cerri, E. Evangelista, P. Cavaliere, Mater. Sci. Eng. A 284 (2000) 254. DOI:10.1016/S0921-5093(00)00748-6 [16] E.J. Zoqui, M.H. Robert: J. Mater. Process Tech. 78 (1998) 198. DOI:10.1016/S0924-0136(97)00483-4 [17] D. Liu, H.V. Atkinson, P. Kapranos: J. Mater. Sci. 39 (2004) 99. DOI:10.1023/B:JMSC.0000007732.04363.81 [18] Y. Birol: J. Alloys Compd. DOI:10.1016/j.jallcom.2009.02.104 [19] Y. Birol: Int. J. Cast Met. Res. DOI:10.1179/174313309X455302. [20] Y. Birol: J. Mater. Process Tech. 173 (2006) 84. DOI:10.1016/j.jmatprotec.2005.09.029 [21] W.F. Miao, D.E. Laughlin: Scripta Mater. 40 (1999) 873. DOI:10.1016/S1359-6462(99)00046-9
443
A
Y. Birol: Effect of solution heat treatment on the age hardening capacity of dendritic and globular AlSi7Mg0.6 alloys
[22] L. Zhuang, R. De Haan, J. Bottema, C.T.W. Lahaye, P. De Smet: Mater. Sci. Forum 331 (2000) 1309. DOI:10.4028/www.scientific.net/MSF.331-337.1309 [23] K. Fukui, M. Takeda, T. Endo: Mater. Lett. 59 (2005) 1444. DOI:10.1016/j.matlet.2004.12.049 [24] K. Matsuda, Y. Sakaguchi, Y. Miyata, Y. Uteni, T. Sato, A. Kamio, S. Ikeno: J. Mater. Sci. 35 (2000) 179. DOI:10.1023/A:1004769305736 [25] C.D. Marioara, S.J. Andersen, J. Jansen: Acta Mater. 51 (2003) 789. DOI:10.1016/S1359-6454(02)00470-6 [26] M. Murayama, K. Hono, M. Saga: Mater. Sci. Eng. A 250 (1998) 127. DOI:10.1016/S0921-5093(98)00548-6 [27] A.K. Gupta, D.J. Lloyd, S.A. Court: Mater. Sci. Eng. A 301 (2001) 140. DOI:10.1016/S0921-5093(00)01814-1 [28] Y. Birol: J. Alloys Compd. 484 (2009) 164 – 167. DOI:10.1016/j.jallcom.2009.05.043. [29] C. Garcia-Cordovilla, E. Louis, J. Narciso, A. Pamies: Mater. Sci. Eng. A 189 (1994) 219. DOI:10.1016/0921-5093(94)90418-9 [30] D.L. Zhang, L.H. Zheng, D.H. StJohn: J. Light Met. 2 (2002) 27. DOI:10.1016/S1471-5317(02)00010-X [31] H. Moller, G. Govender, W.E. Stumpf: Open Materials Science Journal 2 (2008) 6. DOI:10.2174/1874088X00802010006
(Received April 14, 2009; accepted November 4, 2009)
Bibliography DOI 10.3139/146.110293 Int. J. Mat. Res. (formerly Z. Metallkd.) 101 (2010) 3; page 439 – 444 # Carl Hanser Verlag GmbH & Co. KG ISSN 1862-5282
Correspondence address Dr. Yucel Birol Materials Institute, Marmara Research Center, TUBITAK, Gebze, Kocaeli 41470, Turkey Tel.: 90 262 6773084 Fax: 90 262 6412309 E-mail:
[email protected]
You will find the article and additional material by entering the document number MK110293 on our website at www.ijmr.de
W 2010 Carl Hanser Verlag, Munich, Germany
www.ijmr.de
Not for use in internet or intranet sites. Not for electronic distribution.
Applied
444
Int. J. Mat. Res. (formerly Z. Metallkd.) 101 (2010) 3