Effect of hot rolling and heat treatment of Al–5Ti–1B master alloy on the grain refining efficiency of aluminium

Effect of hot rolling and heat treatment of Al–5Ti–1B master alloy on the grain refining efficiency of aluminium

Materials Science and Engineering A301 (2001) 180– 186 www.elsevier.com/locate/msea Effect of hot rolling and heat treatment of Al–5Ti–1B master allo...

882KB Sizes 5 Downloads 80 Views

Materials Science and Engineering A301 (2001) 180– 186 www.elsevier.com/locate/msea

Effect of hot rolling and heat treatment of Al–5Ti–1B master alloy on the grain refining efficiency of aluminium K. Venkateswarlu a, B.S. Murty b,1, M. Chakraborty b,* b

a National Metallurgical Laboratory, Jamshedpur 831007, India Department of Metallurgical and Materials Engineering, Indian Institute of Technology, Kharagpur 721302, India

Received 30 May 2000; received in revised form 14 August 2000

Abstract The influence of hot rolling and annealing of Al–5Ti– 1B master alloy on its grain refining efficiency has been studied in detail. Both hot rolling and annealing improve the grain refining efficiency of the master alloy. Rolling results in the fracture of TiAl3 particles. The amount of deformation required for achieving good grain refinement decreases with increase in rolling temperature. Grain refining efficiency of the master alloy also increases with increase in annealing temperature. The improved grain refining efficiency of the master alloy on annealing is attributed to the increased fraction of TiAl3 and the possible formation of (Ti,Al)B2. © 2001 Elsevier Science B.V. All rights reserved. Keywords: Grain refinement; Heat treatment; Hot rolling; Al– 5Ti– 1B; Aluminium

1. Introduction It is a common industrial practice to add Al –Ti and Al –Ti –B master alloys to molten aluminium alloys prior to casting to achieve fine equiaxed grain structure suppressing the columnar grains. The fine and homogeneous structure in the as cast alloys improves mechanical properties such as fabricability, yield strength and toughness. Most aluminium industries use these master alloys as grain refiners in various compositions. Systematic studies on the effect of Ti:B ratio on the grain refining response of aluminium and its alloys have shown that the Al – 5Ti – 1B master alloy is the best grain refiner [1–4] among all the compositions for most aluminium alloys. It is well known that the addition of Al – Ti – B master alloy to the aluminium melt introduces a large number of intermetallic particles such as TiAl3, TiB2 and (Al,Ti)B2 into the melt [1 – 6]. These intermetallic particles are believed to act as heterogeneous nucleating sites during solidification of aluminium resulting in fine * Corresponding author. Fax: +91-3222-55303. E-mail address: [email protected] (M. Chakraborty). 1 Present address: National Research Institute for Metals, Tsukuba 3050047, Japan.

equiaxed grain structure. Several researchers [1,5 –9] have studied the role of TiAl3 and TiB2 in the master alloys on their grain refining efficiency. The size, size distribution and morphology of TiAl3 particles play an important role in the grain refining efficiency of a given master alloy. The influence of reduction in TiAl3 particle size by thermo-mechanical treatment of as cast master alloy on its grain refining efficiency has not yet been studied in detail. The present paper attempts to fill this gap by a detailed study on the effect of hot rolling and heat treatment of as cast master alloys on their grain refining efficiency.

2. Experimental details The Al –Ti –B master alloys were prepared in an induction furnace by the reaction of molten aluminium (commercial purity) with K2TiF6 and KBF4 salts at 1000°C. The master alloys so prepared and the commercial purity Al were chemically analysed using a flame emission spectrometer (model AA-670 of Shimadzu, Japan), and the compositions are shown in Table 1. The master alloys were hot rolled at 200, 300 and 400°C for reduction levels of 20, 40, 60 and 80%. The rolling process includes a 20% reduction level at

0921-5093/01/$ - see front matter © 2001 Elsevier Science B.V. All rights reserved. PII: S 0 9 2 1 - 5 0 9 3 ( 0 0 ) 0 1 4 2 2 - 2

K. Venkateswarlu et al. / Materials Science and Engineering A301 (2001) 180–186 Table 1 Chemical composition of Al and Ti–B–Al master alloy Alloy

Al Al–Ti–B

the standard linear intercept method using a LECO (DM-400) optical microscope. Scanning electron microscopy was carried out using scanning electron microscopy (model JSM-840A, JEOL, Japan). X-ray diffraction (XRD) patterns of the master alloys were obtained using X-ray diffractometer (model D-500, Siemen’s make).

Composition (wt.%) Fe

Si

Ti

B

Al

0.16 0.16

0.11 0.10

– 5.0

– 1.0

Balance Balance

181

the beginning at a predetermined temperature. For further reduction levels, the samples were again heated to the same temperature and rolled. Similar processing steps were followed for all the rolling experiments. The details of the rolling conditions are given in Table 2. The as cast master alloy was also subjected to heat treatment at 300, 400, 500 and 600°C for 4 h. In order to carry out the grain refinement, 1 kg of commercially pure aluminium was melted in a clay graphite crucible using a pit type resistance furnace under a cover flux; 0.2 wt.% of master alloy (Table 1) was added to the melt at 720°C. The melt was stirred for 30 s immediately on addition of the grain refiner and no further stirring was carried out. Parts of the melt were poured at regular intervals (2, 5, 30, 60 and 120 min) into a cylindrical graphite mould (25 mm diameter and 100 mm height) surrounded by a fire clay brick with its top open for pouring [10]. The cast bars were sectioned at a height of 25 mm from the bottom, and the section surfaces so obtained were polished and etched. Macrostructures were obtained by etching with standard Poulton’s reagent, while etching with Keller’s reagent revealed the microstructures. The microstructural features were studied by optical microscopy, scanning electron microscopy (SEM) and EDS X-ray microanalysis. Grain size analysis was carried out by

3. Results and discussion The X-ray diffraction studies have shown the presence of TiAl3 and TiB2 phase along with a-Al in the as cast Al –Ti –B master alloy. The size of TiAl3 particles was measured and the results of particle size analysis of all hot rolled samples are given in Table 2 along with those of the as cast alloy. The size of 100 particles has been analysed in each case. The results indicate that the mean particle size decreases with the increase in percentage reduction at any rolling temperature, and at a given percentage reduction, it increases with increase in rolling temperature (Fig. 1). The table also gives the standard deviation on the mean particle size, which is quite large, suggesting that particles vary widely in their size. The number of particles below 20 mm also increased with the increase in percentage reduction at all the rolling temperatures (Table 2). The decrease of mean particle size of TiAl3 during hot rolling is due to fracture of these brittle intermetallic particles, which is clearly evident from the SEM studies. Fig. 2a and b show the SEM photomicrographs of as cast master alloy and after 80% reduction at 200°C, respectively. TiAl3 particles (needle/flake shaped) in the as cast condition (Fig. 2a) have become finer in size after

Table 2 Size distribution of TiAl3 particles in as cast and rolled master alloys Master alloy

Processing condition Rolling temperature (°C)

R0 (as cast) R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12

– 200 200 200 200 300 300 300 300 400 400 400 400

Statistical data (mm) Reduction (%)

– 20 40 60 80 20 40 60 80 20 40 60 80

Mean

29.1 23.3 20.1 13.8 10.5 25.2 21.3 18.2 13.8 27.1 24.1 20.3 15.4

S.D.

14 13.3 13.2 9.1 6.3 13.6 13.3 13.0 12.7 13.7 13.2 13.9 12.5

Percent less than 20 mm

Size range Minimum

Maximum

5 5 5 6 4 5 4 4 4 5 4 5 5

70 54 58 46 41 58 52 47 43 63 52 51 49

30 52 59 74 93 36 58 60 72 38 49 59 61

182

K. Venkateswarlu et al. / Materials Science and Engineering A301 (2001) 180–186

Fig. 1. Mean particle size vs. percent reduction of 5/1 master alloys at different temperatures.

the fracture of the TiAl3 particles. As the TiB2 particles are usually very fine ( 0.5 mm) they could not be observed in the present SEM investigation and hence it is difficult to comment on their features during rolling. The fracture of TiAl3 particles is due to stress concentration at the particle matrix interface during rolling. As the particles are brittle, the stress concentration results in their fracture. At lower rolling temperature the stress concentration is expected to be higher and the particles are expected to be more brittle causing extensive fracture of these particles as the rolling temperature is decreased. Fig. 4a–e show the photomacrographs of Al grain refined with as cast master alloy and after rolling at 200°C for different levels of reduction (20 –80% reduction) respectively, and for different holding times. A change from columnar to fine equiaxed grain structure is clearly evident in both the cases with the increase in holding time. However, the grains are much finer in the case of the sample grain refined by the rolled alloy when compared with that refined by as cast master alloy. It is interesting to note that the holding time at which the conversion from columnar to equiaxed grain structure occurs, decreases with increasing amount reduction. This clearly brings out the fact that thermomechanical treatment improves the grain refining efficiency of the master alloys probably due to the fracture of TiAl3 particles to finer sizes on rolling. At the same time the formation of duplex particles due to such treatments cannot be ruled out. This may improve the grain refining efficiency of the hot rolled master alloys. The possibility of formation of duplex particles, (Al,Ti)B2, as a result of heat/thermomechanical treatment exists in view of the observations of Arnberg et al. [2]. These authors had found that TiB2 particles held in liquid Al for a long time show a deficiency in Ti relative to TiB2 stoichiometry and had suggested the formation of duplex particles. Heat treatment or thermomechanical treatment enhances diffusivities that may result in

Fig. 2. Microstructural features of 5/1 master alloys (a) as cast and (b) with deformation.

rolling (Fig. 2b). Fig. 3 shows an intermediate stage (40% reduction at 200°C), which clearly demonstrates

Fig. 3. Fracture of TiAl3 particles in Al– 5Ti– 1B master alloys at 200°C, 40% reduction.

K. Venkateswarlu et al. / Materials Science and Engineering A301 (2001) 180–186

183

Fig. 4. Macrostructures of Al grain refined by Al– 5Ti– 1B master alloys (a) as cast; (b) 20; (c) 40; (d) 60; (e) 80% reduction.

the formation of such duplex particles. However, further investigations are necessary to confirm whether such duplex particles can form at all. On holding the Al melt at 720°C after the addition of the master alloy, the TiAl3 particles start dissolving in the melt, as they are not expected to be stable at the addition level of Ti used for grain refinement. Due to the dissolution, the size of the particles decreases while the Ti content in the melt increases. This may alter the surface tension characteristics of the melt such that heterogeneous nucleation occurs on the large number of sites provided by the remaining TiAl3 and/or TiB2 and/or the duplex particles in the melt. This perhaps

explains the delayed conversion of columnar grain to fine equiaxed grains in the case of the as cast master alloy (Fig. 4a). The holding time at which the conversion from columnar to equiaxed grains occurs is thus related not only to the availability of a sufficient number of fine TiAl3/TiB2/ duplex particles, but also to the surface tension characteristics of the melt in relation to such particles when all other conditions are maintained constant. Hence, with increase in the fraction of fine TiAl3 particles and/or the formation of duplex particles in the hot rolled master alloys, the holding time required for the conversion of columnar to equiaxed grain structure decreases.

184

K. Venkateswarlu et al. / Materials Science and Engineering A301 (2001) 180–186

The grain size analysis of the aluminium grain refined by the master alloy rolled at 200, 300 and 400°C for different amounts of reduction is compared with the as cast master alloy in Fig. 5a – c, respectively. A continuous decrease in the average grain size with the increase of holding time from 2 to 120 min is evident when both as cast and rolled master alloys are used. However, the extent of grain refinement is higher in the case of rolled master alloy when compared with the as cast one at all rolling temperatures. This is direct evidence of the efficacy of fine TiAl3 particles obtained by rolling in improving the grain refining efficiency of the master alloy. Fig. 6a and b show the average grain size of Al grain refined for 2- and 120-min holding with different rolled master alloys. The amount of deformation required to achieve best grain refinement appears to decrease with increase in deformation temperature (Fig. 5). Thus, rolling at 400°C appears to give the best results as good grain refinement could be achieved with a small amount of reduction (40%) of the master alloy. The XRD patterns of the as cast master alloy and after annealing at 300, 400, 500 and 600°C for 4 h are shown in Fig. 7. It is clear from the XRD patterns that the relative peak intensity of TiAl3 increases with heat treatment temperature. Fig. 8 shows the variation of

relative intensity of TiAl3 peak (in percent) with annealing temperature. The relative intensity is the ratio of the intensity of the most intense peak of TiAl3 to that of Al. The increase in the TiAl3 peak intensity with annealing temperature suggests an increase in the volume fraction of the phase. The increase in the TiAl3 fraction on annealing indicates that the alloy does not reach equilibrium during its manufacture, wherein liquid Al is allowed to react with K2TiF6 and KBF4 at 800°C for 60 min. During this reaction, the liquid metal is enriched in Ti from which TiAl3 particles are precipitated. It appears that the precipitation of TiAl3 is not complete during the above reaction and a-Al probably remains in a supersaturated state. Heating the alloy during annealing provides sufficient thermal activation for the diffusion of Ti causing further precipitation or growth of the existing TiAl3 particles. Energy dispersive X-ray microanalysis revealed that the a-Al in as cast and in the heat-treated condition contained 0.36 and 0.24% Ti, respectively. This is therefore, an evidence of the decrease in Ti content on heat-treatment. The larger fraction of TiAl3 observed at higher annealing temperature, may be attributed to faster diffusion. Fig. 9 compares the grain size analysis of Al grain refined with as cast and annealed master alloys. The

Fig. 5. Grain size analysis curves of aluminium, grain refined by Al– 5Ti– 1B master alloys at (a) 200; (b) 300; (c) 400°C.

K. Venkateswarlu et al. / Materials Science and Engineering A301 (2001) 180–186

185

their grain refining efficiency on aluminium. However, with the present understanding it is not clear why the grain refinement is partially lost at higher amounts of reduction when the master alloy is rolled at higher temperatures. It is possible that the TiAl3 particles grow at higher rolling temperatures and higher amount of reduction due to intermediate heating of the alloy between each rolling pass, which could be a possible reason for the loss of grain refining efficiency of the master alloy under these conditions.

Fig. 6. Grain size analysis curves of aluminium, grain refined by different rolled master alloys (Al–5Ti–1B), (a) 2-min and (b) 120-min holding timings.

figure clearly shows that the grain size of Al grain refined with annealed master alloy is lower than that refined with as cast master alloy at all holding times. In addition, the grain size is also finer with increase in annealing temperatures in the range of annealing temperatures studied. The improved grain refining efficiency of the master alloy on annealing could be attributed to the increased fraction of TiAl3. In addition, it is possible that duplex particles (Ti,Al)B2 may also form during annealing which can act as additional nucleating sites as suggested earlier [2]. These particles could contribute partly to the improved grain refining efficiency of the master alloy. Unfortunately, the duplex particles if any, could not be observed in the XRD and SEM studies probably because they are very fine and their volume fraction is quite low. In the light of the results of annealing, the decrease in the amount of deformation required to achieve good grain refinement with increase in rolling temperature can be understood better. It is possible that when the master alloy is rolled at higher temperature, the preheating/intermediate heating between the different reduction levels also influence the volume fraction of TiAl3 as has been seen in the annealing of as cast master alloy. Thus, the heating and rolling both have effects on the volume fraction of TiAl3 particles in the master alloy, which in turn affects

Fig. 7. XRD patterns of Al-5Ti-1B master alloy (a) as cast and 4 h heat-treatment at (b) 300; (c) 400; (d) 500; (e) 600°C.

Fig. 8. Relative percent of TiAl3 vs. different heat-treatment temperatures of Al– 5Ti– 1B master alloys.

186

K. Venkateswarlu et al. / Materials Science and Engineering A301 (2001) 180–186

number of effective nucleating sites for Al. The amount of deformation required to achieve good grain refinement decreases with increase in rolling temperature. Grain refining efficiency of the master alloy improves with increase in annealing temperature. The improved grain refining efficiency of the master alloy on annealing is attributed to the increase in the volume fraction of TiAl3 and the possible formation of duplex particles (Ti,Al)B2.

References

Fig. 9. Grain size analysis curves for different heat-treated Al– 5Ti– 1B master alloys.

4. Conclusions Rolling and annealing of the Al – 5Ti – 1B master alloy improves its grain refining efficiency. Rolling induces fracture of the TiAl3 particles, thus increasing the

.

[1] A. Cibula, J. Inst. Met. 76 (1949– 1950) 321. [2] L. Arnberg, L. Beckrud, H. Klang, Met. Technol. 9 (1982) 7. [3] M.E.J. Birch, P. Fisher, Aluminium Technology, Institute of Metals, London, 1986, p. 117. [4] F.R. Mollard, W.G. Lidman, J.C. Bailey, Light Metals, TMS– AIME, 1987, p. 749. [5] G.P. Jones, J. Pearson, Metall.Trans. 7B (1976) 223. [6] D.G. McCarteny, Int. Mater. Rev. 34 (1989) 247. [7] J. Perason, M.E.J. Birch, J. Met. 31 (11) (1979) 27. [8] P.S. Mohanty, J.E. Gruzleski, Acta Metall. Mater. 43 (1995) 2001. [9] F.A. Grossley, L.F. Mondolfo, Trans. AIME 191 (1951) 1143. [10] C.R. Chakravorty, M. Chakraborty, Cast Met. 4 (1991) 98.