Materials Letters 59 (2005) 3398 – 3401 www.elsevier.com/locate/matlet
Effect of Fe on grain refining of pure aluminum refined by Al5TiB master alloy Yijie Zhang, Haowei Wang *, Naiheng Ma, Xianfeng Li The State Key Laboratory of Metal Matrix Composites, Shanghai Jiaotong University, Shanghai 200030, PR China Received 20 April 2005; accepted 7 June 2005 Available online 1 July 2005
Abstract High pure aluminum and commercial pure aluminum were refined by Al5TiB master alloy, respectively. Experimental results show 0.2 wt.% Al5TiB master alloy can refine commercial pure aluminum effectively, but has no contribution to the refining of high pure aluminum and 1 wt.% Al5TiB master alloy also can refine high pure aluminum effectively. SEM observation indicates that in situ FeAl3 rod phase was found on grain boundary. According to experimental findings, the effect of Fe includes promoting nucleation of primary a-Al and forming FeAl3 rods which can act as pinning effect on grain boundary to hinder growth of a-Al grain. D 2005 Elsevier B.V. All rights reserved. Keywords: Grain refining; Al5TiB master alloy; Aluminum; Scanning electron microscopy (SEM)
1. Introduction Grain refinement has been an important technique for improving the soundness of aluminum products for most of this century . Equiaxed grain structure ensured uniform mechanical properties, reduced ingot cracking, improved feeding to eliminate shrinkage porosity, distribution of second phases and microporosity on a fine scale, improved machinability and cosmetic features. The grain refining innoculants commonly used in the aluminum industry usually are master alloys of Al-Ti or Al-Ti-B . Over the past years, the volume of scientific study on this topic pay more attention on effect of Ti or B [3– 5] and several theories have been proposed though it still contradicted, namely, phase diagram theories, particle theories, the peritectic hulk theory, the hypernucleant theory and the duplex nucleation theory [1,6]. A close study of these theories indicates that all of them are focused on the effect of compound of Ti or B, little interesting in the impurities such as Fe and Si. Some studies considered, to some degree,
* Corresponding author. Tel.: +86 21 62932004; fax: +86 21 62932004. E-mail address: [email protected]
(H. Wang). 0167-577X/$ - see front matter D 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.matlet.2005.06.003
the impurity level will be a small effect and cannot account for main observations . Backerud and Johnsson , Zhiyong Liu et al.  and their co-workers study the effect of silicon on grain refinement of aluminum, they draw the conclusion that in the case of 2% Si content the minimum grain size achieved and grain size increasing with addition of Si content above 3%. In view of Fe is the main source of the impurities for aluminum and its alloys, the purpose of the present investigation is to reveal the influence of Fe on the grain refinement of aluminum by studying commercial pure aluminum (CPAl) and high pure aluminum (HPAl) refined by Al5TiB master alloy. Basing on the experimental results the mechanism of grain refinement was discussed.
2. Experimental procedure Al5TiB master alloy self-prepared was employed for studying grain refinement. For studying the effect of Fe, CPAl (compositions given in Table 1) and HPAl (> 99.99 wt.%Al) were used for the present purpose. After CPAl and HPAl had been melted in different graphite crucibles by resistance furnace, 0.2 and 1 wt.%
Y. Zhang et al. / Materials Letters 59 (2005) 3398 – 3401 Table 1 Chemical composition of CPAl Material
Composition (wt.%) Fe
Al5TiB master alloy was added into the melts, respectively. In each case the melt was maintained at 725 T 5 -C and slow stirred needed to disperse Al5TiB master alloy uniformity in the melt for about 30 s with a graphite rod after addition was completed melting. The melt was held for 10 min and then cast into chill T moulds of Standard Renault Golf which was held at the room temperature. Each cone is sectioned perpendicular to its axis 51mm from its base for macro-analysis, ground mechanically on 200 grit and 400 grit water-proof adhesive papers and then etched in reagent (60 mL HCl, 30 mL HNO3, 5 mL HF, 5 mL H2O) to reveal the macrostructural feature. A Field Emission Scanning Electron Microscope (FESEM) fitted with a link energy dispersive spectrometer (EDS) was employed to study the grain boundaries in this research. Point analysis of EDS was used to determine phase compositions observed on grain boundaries. The samples for SEM analysis were etched in 0.5% HF solution for about 2 min to highlight the grain boundaries.
3. Results and discussion Fig. 1 shows macrographs of CPAl before and after grain refinement. It can be seen clearly from Fig. 1(a) that in the absence of an addition of Al5TiB grain refiner, the CPAl shows complete coarse columnar grain structure, its grain size is about 1mm, while Fig. 1 (b), CPAl with 0.2 wt.% Al5TiB master alloy addition, exhibits fine equiaxed grains, its grain size is about 130 um.
Fig. 2 indicates macrographs of HPAl with different addition of Al5TiB grain refiner. In sharp contrast, Fig. 2 (a) shows that grain refining was not observed in the case of 0.2 wt.% Al5TiB master alloy addition, coarse columnar grains were still obtained and only a little thin layer of coarse equiaxed grains was observed at the periphery of the sample. Fig. 2 (b) exhibits fine equiaxed grains only were obtained, its grain size is about 80 um, in this case 1 wt.% Al5TiB master alloy was added into HPAl. As expected, Al5TiB master alloy can refine CPAl effectively with the addition of 0.2 wt.% into CPAl. But 0.2 wt.% Al5TiB master alloy has no contribution on HPAl grain refining. When Al5TiB master alloy content ups to 1 wt.%, fine equiaxed grains also were obtained for HPAl. In other words, only content of excesses Ti is enough in melt can Al5TiB master alloy refine HPAl effectively. A careful study of CPAl and HPAl indicates that the differences of these two materials are mainly on chemical composition, exactly to say, on content of element Fe. Experiment results indicate that though Fe content is somewhat very little, this is enough for CPAl grain refining. The grain refining includes nucleation and growth. To explain the influence of Fe on grain refining, the growth restriction factor (GRF) was suggested to represent the effect of solute concentration, such as Ti, Fe, on grain size and crystal morphology. The grain size will reach a minimum at a critical GRF . Within certain values, the bigger GRF value, the stronger capacity of promoting nucleation is. The number of foreign clusters for nucleation sites will increase and the surface tension of the liquid aluminum might possibly decrease, so the nucleation rate might increase with Fe existence. The results that 0.2 wt.% Al5TiB master alloy can refine CPAl and cannot refine HPAl maybe due to existence of Fe increases GRF to critical value. This is why the best results were obtained in the case of 0.2 wt.% Al5TiB added into CPAl and 1 wt.% Al5TiB added into HPAl. For CPAl without Al5TiB master alloy addition, the sole effect of Fe is not enough to promote nucleation, so macrograph of this material is still coarse columnar grains. Following the macroscopy analysis of grain size on refined aluminum, a thorough microstructural evaluation of the aluminum was undertaken using FE-SEM and EDS microanalysis to understand the various features observed. Fig. 3 was gained
Fig. 1. Macrographs of CPAl without refinement (a), refined with 0.2 wt.% Al5TiB master alloy (b).
Y. Zhang et al. / Materials Letters 59 (2005) 3398 – 3401
Fig. 2. Macrographs of HPAl refined with 0.2 wt.% Al5TiB master alloy (a), with 1 wt.% Al5TiB master alloy (b).
from grain boundary of CPAl sample with 0.2 wt.% Al5TiB master alloy addition. Interestingly, phases in shape of rod on grain boundary were found and its diameter is about 250 nm in size, in this case which can be called in situ rod phase. EDS was employed herein to confirm compositions of rod phases. EDS result shows that the in situ rod phases are Al – Fe intermetallics, its composition includes 6 wt.% Fe and 94 wt.% Al. In present study this composition cannot be quantified as ingredient of in situ Al – Fe phase due to field for data acquisition on sample surface by electron spot is about 1um, which is bigger than rods diameter. Thus EDS result is not accuracy for calculating weight ratio of Al and Fe. In addition, element Fe was not detected within primary a-Al by EDS analysis. According to the Al – Fe phase diagram , seen in Fig. 4, the solubility of Fe in aluminum at 655 -C is about 0.04 wt.% under equilibrium solidification condition, and Fe content at the eutectic point is about 1.8 wt.%. In present investigation CPAl contains 0.14 wt.% Fe, basing on phase diagram, in situ rod phases on grain boundary are FeAl3 phase, which can subsequently hinder the grain growth during solidification. In general, in the case of 0.2 wt.% Al5TiB added into CPAl, the effect of Fe promotes nucleation of primary a-Al and forms FeAl3
rods acting as pinning effect on grain boundary to hinder growth of a-Al grain.
4. Conclusions In the present study, the effect of Fe on grain refinement of pure aluminum has been investigated by studying CPAl and HPAl refined by Al5TiB master alloy. The major conclusions and suggestions drawn from the results are as follows: CPAl can be refined effectively by 0.2 wt.% Al5TiB master alloy, but 0.2 wt.% Al5TiB master alloy has no contribution on grain refining of HPAl. Also 1 wt.% Al5TiB master alloy can refine HPAl effectively. FeAl3 rods were observed on grain boundary of CPAl sample with 0.2 wt.% Al5TiB master alloy addition. Its diameter is about 250 nm in size. The existence of Fe plays an important role on grain refining of CPAl, its influence on grain refining may have two different kinds of effects, that is, solute Fe atoms promote nucleation of primary a-Al crystal during solidification process and in situ FeAl3 rods can act as pinning effect on grain boundary and hinder grain growth. Liq. + Al 727
-1.8% 655°C 0.04%
Al Fig. 3. SEM observation on grain boundary of CPAl sample with 0.2 wt.% Al5TiB master alloy addition.
Fig. 4. The aluminum end of the aluminum – iron equilibrium diagram.
Y. Zhang et al. / Materials Letters 59 (2005) 3398 – 3401
References  Mark Easton, David Stjohn, Metallurgical and Materials Transactions 30A (1999) 1613.  P.S. Mohanty, J.E. Gruzleski, Acta Metallurgica Et Materialia 43 (1995) 2001.  M.M. Guzowski, G.K. Sigworth, D.A. Sentner, Metallurgical Transactions. A, Physical Metallurgy and Materials Science 18A (1987) 603.  L. Backerud, Shao Yidong, Aluminum 67 (1991) 780.  C. Limmaneevichitr, W. Eidhed, Materials science & engineering. A, Structural materials: properties, microstructure and processing 349 (2003) 197.
 A.L. Greer, A.M. Bunn, A. Tronche, et al., Acta Materialia 48 (2000) 2823.  M. Johnsson, L. Backerud, G.K. Sigworth, Metallurgical Transactions. A, Physical Metallurgy And Materials Science 24 (1993) 481.  L. Backerud, M. Johnsson, Light Metals 27 (1996) 679.  Zhiyong Liu, Mingxing Wang, Yonggang Weng, et al., Materials Transactions 44 (2003) 2157.  L.F. Mondolfo, Aluminum Alloys: Structure and Properties, Butworth and Co Ltd., 1976.