Modification of Hypereutectic Al–20 wt%Si Alloy Based on the Addition of Yttrium and Al–5Ti–1B Modifiers Mixing Melt

Abstract

The effect of the addition of yttrium (Y) and Al–5Ti–1B modifiers mixing melt (3M) on the microstructure and mechanical properties of hypereutectic Al–20Si alloy was studied in the present work. The as-cast specimens were examined by scanning electron microscope equipped with energy spectrometer, electron probe microanalysis and X-ray diffraction. The results demonstrated that the morphology of primary Si could be refined from coarse irregular star-like/plate-like structures into fine blocks with the addition of 0.6 wt% Y and 1.0 wt% Al–5Ti–1B at 650 °C. The average grain size of primary Si was reduced from 82 to 29 µm, and the aspect ratio was decreased from 1.81 to 1.47. Similarly, the eutectic Si structure was modified from coarse needlelike/flake-like structures into fine coral fibrous structures and partial granular structures, with the mean roundness of eutectic Si decreased from 7.8 to 2.32. In addition, the coarse α-Al dendrites were significantly refined into the uniform equiaxed dendrites. With the refinement and homogenization of Si phases, the optimal modified alloy obtained the optimal ultimate tensile strength (UTS) and elongation (EL). The UTS was enhanced from 94 to 154 MPa, and the EL was increased from initial 1.12 to 1.79%. Furthermore, the refinement and modification mechanism of the addition of yttrium and Al–5Ti–1B modifiers mixing melt on Al–20Si alloy were also discussed.

Technical Review and Discussion

Addendum: Observations on Proof by IJMC Advisory and Board of Review Member Dr. Geoffrey K Sigworth and Answers from the Authors

Observation Historically both ‘modification’ and ‘refinement’ have been used to denote the changes in silicon structure in hypereutectic Al–Si alloys. This lack of clarity in terminology is particularly confusing in this paper, because both an yttrium addition and an Al–Ti–B grain refiner were employed. To make matters clear, the term ‘modification’ should be used to denote changes in the eutectic silicon; and the term ‘refinement’ should refer only to the reduction in size of the primary phase (silicon in hypereutectic, and aluminum in hypoeutectic alloys).

Answer Special thanks for your good comment. I am very sorry for the confusion caused by my inappropriate description. As Geoffrey K Sigworth said, actually, the “refinement” and “modification” are two different concepts. The refinement mainly refers to the change of the size of the primary phase, for instance, the reduction in the size of primary Si in hypereutectic Al–Si alloy. However, the modification mainly indicates the change of eutectic Si in Al–Si alloy. Therefore, we have carefully examined and checked the description of both primary Si and eutectic Si. The final paper has been revised. We have correctly used the “refinement” to describe the reduction in size of the primary Si, and used the “modification” to state the change of the eutectic Si.

Observation Normally one would not add a grain refiner to a hypereutectic alloy, since the primary phase is silicon, not aluminum. However, the large yttrium additions slowed silicon growth sufficiently, so that large dendritic grains of aluminum also formed (see Fig. 16a). Some of these aluminum dendrites were large, so Al–5Ti–1B refiner was also used. I have never seen aluminum dendritic grains like this in a hypereutectic alloy. Normally the aluminum phase is present as small ‘globules,’ or sometimes as a ‘halo’ around primary silicon. A typical structure in 390 alloy (containing 16.4% Si) is shown in Fig. 16b.47

Figure 16

(a) SEM image of aluminum dendrites in alloy A2 (Fig. 11a in the paper) and (b) optical micrograph of sample cut from the dome section of a 390 Alloy piston (refined with 45 ppm P).47

Answer Special thanks to your good comment. In order to obviously demonstrate the change of α-Al dendrite, the corresponding SEM image of α-Al dendrite with high magnification (300 × ) is shown in Fig. 11a in the paper. However, the partially microstructural description lead to a misunderstanding about the refinement of eutectic α-Al phase. According to our experiments and research, the distribution of eutectic α-Al phase in Al-Si alloy with the high content of silicon has two different situations. On the one hand, the α-Al phase grows into the small ‘globules’, or sometimes as a ‘halo’ around primary Si, as shown by the red circle in Fig. 17a. On the other hand, the α-Al dendrite will be formed in the areas where primary silicon is sparse, as shown the red rectangle in Fig. 17a. The similar situation also appears in 390 alloy refined with 45 ppm P, the corresponding areas marked with the red circle and rectangle in Fig. 17b, respectively.

Figure 17

(a) Low-magnification SEM im age of aluminum dendrites in alloy A2 (Fig. 5b in the paper) and (b) optical micrograph of sample cut from the dome section of a 390 Alloy piston (refined with 45 ppm P).47

According to the reports (Refs. [1, 43] in the paper), the addition of Ge and Ni–Si inoculation could improve the temperature of eutectic reaction and shorten two-phase coexistence area. Therefore, the nucleation temperature of α-Al and eutectic Si will be reduced. Similarly, the addition of rare earth yttrium also could increase the nucleation efficiency of α-Al and eutectic Si. The α-Al phase that is formed around the primary Si always presents the small ‘globules’, or sometimes as a ‘halo’ around primary Si, which is mainly attributed to the α-Al phase normal growth attached to primary Si. Meanwhile, the primary Si can be refined because the rare earth atoms are absorbed in the groove of the twin crystal and inhibits the growth of primary Si. However, the eutectic α-Al phase that is formed in the areas where the primary Si is rare and uneven distribution will quickly grow and coarsen during the solidification process. Therefore, we added the Al–5Ti–1B modifier into the Al–20Si alloy melt to change the morphology of α-Al dendrite. From Fig. 12 in the paper, it can be clearly seen that the TiB₂ particles act as the nucleation substrate of α-Al to enhance the nucleation rate, and the (TiY) cluster further restricts the growth of α-Al dendrite. The specific results are shown in Fig. 11 in the paper. However, we have done no further study for the detailed formation mechanism of the α-Al phase with different situations in the paper. According to this review advice, we will conduct a detailed study on the formation process and mechanism of this phenomenon.

Thank you again for your detailed and valuable comments, we will do more detailed research work based on your advice.

Observation The Tibor addition employed in this study was one part in a hundred. The normal addition in hypoeutectic alloys is 1/10 to 1/5 of that. The yttrium additions were also quite large. Hence, the proposed process will be rather costly. It is not clear that improvements in cast structure will justify the additional cost. The refinement of primary silicon was modest—from 82 to 29 microns. In 390 alloy, a good refinement with phosphorus easily produces an average primary silicon size of 25 microns or less (Figs. 16b, 17b, 18).

Figure 18

Effect of mold and casting temperature on primary silicon size on ‘piston’ castings of 390 alloy containing 45 ppm P. Note: Results are shown for three casting temperatures: 725, 775 and 825 C. (Reproduced from [1]).47

Answer Special thanks to your good comment. As Professor G. K. Sigworth said, a good refinement with phosphorus easily produces an average primary silicon size of 25 microns or less in 390 alloy. The research has been confirmed that the phosphorus has a pleasant refinement effect on primary Si because the Al₃P particles can act as heterogeneous nucleation core of primary Si to enhance nucleation efficiency. Comparing with Si crystals, the high melting point (about 1060 °C) Al₃P compound encompasses identical crystals structure (FCC), nearby lattice constant (a_Al3P = 0.546 nm, a_Si = 0.54 nm) and low lattice mismatch. Unfortunately, the phosphorus addition can only effectively refine primary silicon. The reports (Ref. [40] in the paper) proposed sodium and strontium can modify the eutectic Si because Na and Sr own an appropriate atomic radius ratio (~ 1.65) with Si atom, which promote the generation of twins and change the growth direction of eutectic Si. However, the modification ability will disappear due to producing the poison phenomenon when P and Na/Sr are simultaneously added into the hypereutectic Al-Si alloys (Ref. [33] in the paper). Our previous research (Ref. [9] in the paper) has confirmed that rare earth yttrium can not only refine primary Si from irregular morphology to small block-like, but also modify the eutectic Si from acicular to fibrous. In the present work, we added the rare earth yttrium and Al–5Ti–1B modifiers mixing melt into the hypereutectic Al-Si alloy melt to simultaneously refine primary Si and modify eutectic structure (α-Al and eutectic Si). In order to reduce energy consumption, the rare earth yttrium is introduced in the form of Al–20Y master alloy with the 200 RMB/kg. The cost of Al–5Ti–1B modifier and Al–5P master alloy is 40 and 30 RMB/kg, respectively. The detailed cost, advantage and disadvantage of P or Y/(Al-5Ti-1B) addition is shown in the table below (take 100 g Al–20Si or 390 alloy as an example):

Modifier	Cost /RMB	Advantage	Disadvantage
P (45 ppm)	(100 × 0.000045 ÷ 0.05) × 0.03 = 0.0027	Good refinement effect on primary Si	Need high temperature, high energy consumption
Y/(Al–5Ti–1B) (0.6%/1.0%)	(100 × 0.6% ÷ 0.2) × 0.2 + (100 × 1.0%) × 0.04 = 0.64	Simultaneously refine primary Si and modify eutectic structure, green and energy saving	Relatively higher cost

Through the above comparative analysis, it can be accepted that the effect of the addition of yttrium and Al–5Ti–1B modifiers mixing melt on microstructure evolution of Al–20Si alloy has some scientific research value. Furthermore, we will continue to improve the modification process to reduce costs and increase efficiency according to Dr. G. K. Sigworth’s great and valuable advice.

Observation Tensile properties were measured in this study. But these alloys are used for wear resistance, and refinement of primary silicon is done to improve machinability. The latter is most important. In conventional (hypoeutectic) alloys, the cost of machining an engine block is often greater than the cost of the raw casting. Machining costs are greater in 390 and other hypereutectic alloys, since diamond cutting tools and special surface treatments are needed. Neither machinability nor wear resistance were studied, so it is not clear if the proposed process will have any commercial impact.

Answer Special thanks to your good comment. The question raised by Dr. G. K. Sigworth also is exactly what we want to consider next. Although the Al–Si alloy has better wear resistance, the machinability of the alloy decreases with the increasing of Si content, which makes the subsequent machining cost as expensive. Therefore, more and more scholars paid their attention to research the multicomponent near-eutectic casting Al-Si piston alloy (Refs. [2, 32, 39] in the paper). The traditional alloying element is Cu, Ni and Mg in near-eutectic Al–Si alloy. In the present work, we firstly introduced the yttrium and Al–5Ti–1B modifiers mixing melt into the hypereutectic Al–20Si alloy to obtain the fine structure and enough ultimate tensile strength for manufacture of the engine block or piston. In our next work, we have studied the effect of yttrium addition on the microstructure and mechanical properties of multicomponent near-eutectic Al–Si–Cu–Ni–Mg piston alloy. According to Dr. G. K. Sigworth’s valuable advice, we will investigate the wear resistance and machinability of multicomponent Al-Si piston alloy. In summary, our purpose is to establish a relatively complete system to study the refinement and modification effect of rare earth elements in Al–Si piston alloy.

Thank you again for your good and valuable comments, we will do more detailed research work based on your advice.

We would like to express our sincere thanks to the Editor-in-Chief Thomas Prucha and Dr. Geoffrey K Sigworth for the constructive and positive comments.