Si poisoning and promotion on the microstructure and mechanical properties of Al–Si–Mg cast alloys

Abstract

Hypoeutectic Al–xSi–0.45Mg (x = 6.5, 7.5, 8.5, 9.5) alloys were refined by Al5Ti1B containing TiB₂ and Al3Ti3B containing TiB₂ and AlB₂, respectively. With increasing Si, Si poisoning on TiB₂ results in the obvious coarsening of primary α-Al in Al5Ti1B-refined alloys from 350 ± 40 to 400 ± 50, 475 ± 50 and 560 ± 80 μm, and the competition between Si promotion on AlB₂ and Si poisoning on TiB₂ leads to the slight coarsening of primary α-Al in Al3Ti3B-refined alloys from 215 ± 30 to 265 ± 35, 265 ± 30 and 315 ± 25 μm. After T6 heat treatment, with increasing Si, the yield strength (YS) of Al5Ti1B-refined alloys increases from 294 ± 2 to 299 ± 2, 304 ± 1 and 309 ± 2 MPa, and the elongation first increases from 3.5 ± 0.8 to 4.5 ± 1.0 and 7.8 ± 1.4%, after decreases to 5.5 ± 1.2%, while the YS of the Al3Ti3B-refined alloys increases from 300 ± 1 to 305 ± 2, 312 ± 1 and 317 ± 2 MPa, and the elongation increases from 6.1 ± 1.1 to 8.5 ± 1.2, 11.8 ± 1.5 and 12.1 ± 1.6%. The increase in the secondary phase and precipitation strengthening results in the increase in strength with increasing Si. With increasing Si, the decrease in porosity formation by decreasing solidification interval and increasing fluidity is superior to the increase in porosity formation by slightly coarsening grain size, which leads to the increase in ductility in the Al3Ti3B-refined alloys, while the competition between porosity decreasing and increasing factors leads to the inverted ‘V’-shaped evolution of ductility in the Al5Ti1B-refined alloys.

Introduction

Al–Si–Mg cast alloys have been widely used for making high-integrity castings with a combination of good castability, low density, high-strength-to-weight ratio, good corrosion resistance and low coefficient of thermal expansion, which are necessary for transport manufacturing to provide light-weighting components. Grain refinement has been proved as an important way to obtain fine primary α-Al grains, which can improve the toughness, strength, formability and machinability [1,2,3,4,5,6,7,8,9,10,11].

The most widely used grain refiner in aluminium alloys over the past several decades is the Al5Ti1B master alloy with TiB₂ particles and excess Ti, which inoculates the melt with TiB₂/TiAl₃ particles as heterogeneous nuclei, and the sufficient free Ti solute in the melt can restrict the growth of primary α-Al grains after nucleation. The exact mechanism of grain refinement under Al5Ti1B has been well demonstrated recently; the formation of a monolayer of (112) Al₃Ti two-dimensional compound on the (0001) TiB₂ surface can reduce the misfit between TiB₂ and α-Al from − 4.2 to 0.09%, which can efficiently enhance the nucleation potency of TiB₂ particles for primary α-Al grains [12]. The Al5Ti1B master alloy offers good performance in the casting of wrought alloys, but it is hard to meet the expectations in cast Al–Si alloys, especially with a content of Si higher than 3.5 wt% [13,14,15]. The reason is that Si in the melt reacts with Ti to form Ti–Si phases, which poison the TiB₂ nucleation site [16, 17]. The poisoning mechanism of Si on the TiB₂ nucleation sites has been verified by experiment recently, and Al–Si–Ti particles were observed on the prism face of TiB₂ particles [18].

To reduce or avoid Si poisoning, one effective way is to introduce B into the Al–Si alloys, and it is often achieved by adding Al–B master alloy, in which AlB₂ resides as the source to supply B, into the melt. The use of Al–B master alloys for the grain refinement of Al–Si cast alloys dated back to 1980s, and the grain size was continually reduced even with increasing Si content [15]. AlB₂ is the dominated particle in Al–B master alloys and has a small misfit between α-Al, and it was expected that AlB₂ could be a potent nucleating substrate for α-Al [19]. However, a number of observations [20, 21] showed that AlB₂ alone without Si cannot effectively refine α-Al, indicating that the solute Si may interfere with AlB₂ to enhance its nucleating potential. There is still lack of unambiguous understanding of the mechanism. Recently, it has been proposed that the creation of a layer of SiB₆ at the interface between AlB₂ and Al may reduce the crystallographic mismatch, which can significantly improve the nucleating potency of AlB₂, and the enhanced grain refining efficiency can be mainly attributed to the enhanced heterogeneous nucleation of AlB₂ caused by Si [22]. However, an amount of ~ 0.12 wt% Ti is usually present in commercial cast aluminium alloys for grain growth restriction. It was found that the commercial Al–Si–Mg cast alloys could not enjoy the outstanding grain refinement efficiency of Al–B master alloys, with the presence of Ti, since the AlB₂ particles are readily transformed to TiB₂ particles and then suffering from the Si poisoning, and the grain refinement efficiency of Al–B master alloys is much similar to Al5Ti1B master alloy for commercial Al–Si–Mg cast alloys with Ti present [23].

For the grain refinement of commercial Al–Si–Mg cast alloys containing Ti, one effective way is to reduce the Ti content and increase the B content in the Al–Ti–B master alloys, and Al3Ti3B master alloy with TiB₂ particles and excess B was found providing effective grain refinement [10, 11, 23, 24]. Al3Ti3B master alloy was reported containing TiB₂ and AlB₂ particles [10, 11, 23], and both of these particles could be potentially heterogeneous nucleation sites. The exact mechanism of grain refinement under Al3Ti3B is still quite unclear. On the one hand, Si poisons the heterogeneous nucleation of TiB₂ particles; on the other hand, Si promotes the heterogeneous nucleation of AlB₂ particles, so it is interesting to study the effect of Si poisoning and promotion on the grain refinement of Al–Si–Mg cast alloys under Al3Ti3B. Furthermore, seldom did research focus on Si poisoning and promotion on the mechanical properties of Al–Si–Mg cast alloys.

The objective of this paper is to study the multiple effects of Si on the microstructure and mechanical properties of hypoeutectic Al–Si–Mg cast alloys under different grain refiners, especially Si poisoning under Al5Ti1B and Si promotion under Al3Ti3B, to provide high-performance cast Al–Si–Mg alloys with high strength and high ductility and meet the increasing requirements in automotive industry.

Experimental

Materials and melt preparation

A serial of hypoeutectic Al–Si–Mg cast alloys with 0.45 wt% Mg and different Si contents (6.5, 7.5, 8.5 and 9.5 wt%) were prepared and melted in 12-kg capacity clay–graphite crucibles separately using the electric resistance furnace, and the detail compositions of the investigated alloys were measured by inductively coupled plasma atomic emission spectroscopy (ICP-AES) and are listed in Table 1. During melting, the temperature of the furnace was controlled at 750 °C. After 1 h of homogenisation, Al–10 wt% Sr master alloy was added into the melt to make the desired Sr content of 140 ppm for modification. The melt was subsequently degassed through injecting pure argon into the melt by using a rotary degassing impeller at a speed of 350 rpm for 4 min. After degassing, the melt was hold for 10 min for temperature recovery, followed by adding 0.2 wt% Al5Ti1B or 0.2 wt% Al3Ti3B for grain refinement.

Table 1 Chemical compositions of experimental alloys analysed by ICP-AES (wt%)

Casting process and heat treatment

With the intention of casting tensile test bars, the prepared melt was poured at 720 °C into an ASTM B-108 permanent mould preheated at 460 °C, as shown in Fig. 1a. Figure 1b shows the gravity casting made by the permanent mould, and two round tensile test bars were made from each casting, as indicated by the dashed rectangle box in Fig. 1a. With the intention of testing the fluidity of the investigated alloys, the prepared melt was poured at 720 °C into an ASTM standard spiral flow fluidity test mould preheated at 460 °C. Three fluidity tests and three density tests were made for each alloy to give the average spiral flow length and average porosity percentage with error bar, respectively. The cast tensile test bars were subjected to T6 heat treatment, including solution treatment and artificial ageing. Solution treatment was carried out at 540 °C for 8 h, followed by immediate water quenching to room temperature. Ageing treatment was performed at 170 °C for 8 h, followed by air cooling to room temperature.

Figure 1

a Permanent mould made according to ASTM B–108, and b key dimensions of the gravity casting tensile test bar made by the mould

Microstructure characterisation and tensile tests

The microstructure was examined using the Zeiss optical microscopy (OM), the Zeiss scanning electron microscope (SEM), the JEOL-2100 transmission electron microscopy (TEM) and the D8 X-ray diffraction (XRD) instrument. The specimens for OM and SEM analysis were prepared by the standard technique of grinding. Polarised OM observation of grain size was performed after anodised with Barker solution (97 vol% H₂O and 3 vol% HBF₄). SEM analysis was conducted after etching with the Keller solution (1 vol% HF, 1.5 vol% HCl, 2.5 vol% HNO₃ and 95 vol% H₂O). Five polarised OM images with a magnification of 25 were counted to give each of the statistical average grain sizes with error bar. Thin specimens for TEM observation were prepared by standard electropolishing. The electrolytic solution was a mixture of nitric acid and methyl alcohol (2:8), used at − 20 to – 30 °C and 20 V. TEM operating at 200 kV was used for bright-field imaging and high-resolution TEM (HRTEM) imaging. XRD analysis was conducted from 2θ degrees 25°–90°. Tensile tests were conducted at room temperature following the ASTM B557 standard using an Instron 5500 Testing System. Each tensile test data reported with error bar were based on the mechanical properties obtained from 6 to 8 samples.

Results

As-cast microstructure

Figure 2a–d presents the polarised optical micrographs showing the grain size of primary α-Al in the as-cast Al–xSi–0.45Mg alloys with 6.5 wt% Si, 7.5 wt% Si, 8.5 wt% Si and 9.5 wt% Si, respectively, under the refinement of Al5Ti1B. When the Si content is increased to 7.5 wt%, with the increase in Si, the grain size of primary α-Al phase is coarsened significantly, which indicates that the poisoning of Si on grain refinement is significant when the Si content is up to 7.5 wt%, under the refinement of Al5Ti1B.

Figure 2

Polarised optical micrographs showing the grain size of primary α-Al phase in the as-cast Al–xSi–0.45Mg alloys refined by Al5Ti1B: a 6.5 wt% Si, b 7.5 wt% Si, c 8.5 wt% Si and d 9.5 wt% Si

Figure 3a–d presents the polarised optical micrographs showing the grain size of primary α-Al phase in the as-cast Al–xSi–0.45Mg alloys with 6.5 wt% Si, 7.5 wt% Si, 8.5 wt% Si and 9.5 wt% Si, separately, under the refinement of Al3Ti3B. The grain size of primary α-Al phase in the Al–xSi–0.45Mg alloys refined by Al3Ti3B is obviously smaller than that of the alloys refined by Al5Ti1B. With the increase in Si, the coarsening of the primary α-Al in the Al–xSi–0.45Mg alloys refined by Al3Ti3B is not obvious, which indicates that the poisoning of Si on the grain refinement of Al–xSi–0.45Mg alloys is weak with the Si content up to 9.5 wt%, under the refinement of Al3Ti3B.

Figure 3

Polarised optical micrographs showing the grain size of primary α-Al phase in the as-cast Al–xSi–0.45Mg alloys refined by Al3Ti3B: a 6.5 wt% Si, b 7.5 wt% Si, c 8.5 wt% Si and d 9.5 wt% Si

Figure 4 shows the statistical average grain size of the primary α-Al phase in the as-cast Al–xSi–0.45Mg (x = 6.5, 7.5, 8.5, 9.5) alloys refined by Al5Ti1B and Al3Ti3B. Under the refinement of Al5Ti1B, the grain size of the primary α-Al phase is 350 ± 40 μm with a Si content of 6.5 wt%; with the increase in Si content to 7.5, 8.5 and 9.5 wt%, the primary α-Al phase is coarsened obviously to 400 ± 50, 475 ± 50 and 560 ± 80 μm. The grain size of the primary α-Al phase increases with the increase in Si content when refined by Al5Ti1B; the grain size of the primary α-Al phase is coarsened obviously when the Si content is up to 7.5 wt% and coarsened nearly linear after. Under the refinement of Al3Ti3B, the grain size of the primary α-Al phase is fine as 215 ± 30 μm at 6.5 wt% Si, and the grain size is increased to 265 ± 35 μm at 7.5 wt% Si, then the grain size is maintained at 265 ± 30 μm at 8.5 wt% Si, after the grain size is increased to 315 ± 25 μm at 9.5 wt% Si. The grain size of the primary α-Al phase in the Al3Ti3B-refined alloy is significantly smaller than that of the Al5Ti1B-refined alloy. With the increase in Si, the coarsening of the primary α-Al phase in the Al3Ti3B-refined alloy is obviously slighter than that of the Al5Ti1B-refined alloy.

Figure 4

Statistical average grain size of primary α-Al phase in the as-cast Al–xSi–0.45Mg (x = 6.5, 7.5, 8.5, 9.5) alloys refined by Al5Ti1B and Al3Ti3B

Figure 5a–d shows the SEM morphology of the as-cast hypoeutectic Al–xSi–0.45Mg alloys with 6.5 wt% Si, 7.5 wt% Si, 8.5 wt% Si and 9.5 wt% Si, respectively, under the refinement of Al5Ti1B master alloy. Figure 5e–h shows the SEM morphology of the as-cast Al–xSi–0.45Mg alloys with 6.5 wt% Si, 7.5 wt% Si, 8.5 wt% Si and 9.5 wt% Si, separately, under the refinement of Al3Ti3B master alloy. The insert in each figure shows the SEM morphology with high magnification. Primary α-Al phase, eutectic Si phase and β-Mg₂Si intermetallic phase coexist in the as-cast alloys refined by both Al5Ti1B and Al3Ti3B. β-Mg₂Si phase is located in the Al–Si eutectic region. With the increase in Si, the fraction of eutectic Si phase in the as-cast Al–xSi–0.45Mg alloys increases, for the condition both refined by Al5Ti1B and Al3Ti3B.

Figure 5

SEM micrographs showing the morphology of the as-cast Al–xSi–0.45Mg alloys refined by: a–d Al5Ti1B and e–h Al3Ti3B with a, e 6.5 wt% Si, b, f 7.5 wt% Si, c, g 8.5 wt% Si and d, h 9.5 wt% Si

Microstructure after heat treatment

Solution treatment can spheroidise the eutectic Si phase and dissolve intermetallic phases to form saturated solid solution [25]. Figure 6a–d shows the SEM morphology of the T6 heat-treated Al–xSi–0.45Mg alloys with 6.5 wt% Si, 7.5 wt% Si, 8.5 wt% Si and 9.5 wt% Si, respectively, under the refinement of Al5Ti1B. Figure 6e–h shows the SEM morphology of the T6 heat-treated Al–xSi–0.45Mg alloys with 6.5 wt% Si, 7.5 wt% Si, 8.5 wt% Si and 9.5 wt% Si, separately, under the refinement of Al3Ti3B. The insert in each figure shows the SEM morphology with high magnification. Eutectic Si phase is spheroidal morphology, which indicates that the eutectic Si phase is spheroidised after T6 heat treatment. The spheroidised Si particles are fine, which are beneficial to ductility [26]. The morphology of the spheroidised Si particles in the Al5Ti1B- and Al3Ti3B-refined alloys is much similar. No β-Mg₂Si intermetallic phase was observed, which indicated that the β-Mg₂Si phase was well dissolved into the α-Al matrix after the solution treatment. The well solid solution of β-Mg₂Si phase could ensure the precipitation of nanoscale strengthening precipitates in the α-Al matrix after ageing treatment, which contributes to the strengthening of the alloys after T6 heat treatment. With the increase in Si, the volume fraction of spheroidised Si phase increases, for the condition refined by both Al5Ti1B and Al3Ti3B.

Figure 6

SEM micrographs showing the morphology of the T6 heat-treated Al–xSi–0.45Mg alloys refined by a–d Al5Ti1B and e–h Al3Ti3B with a, e 6.5 wt% Si, b, f 7.5 wt% Si, c, g 8.5 wt% Si and d, h 9.5 wt% Si

Figure 7a–d presents the bright-field TEM micrographs showing the β″ strengthening precipitate in the Al–xSi–0.45Mg alloys with 6.5 wt% Si, 7.5 wt% Si, 8.5 wt% Si and 9.5 wt% Si, respectively, after T6 heat treatment. Embedded and lying β″ precipitates were found in the α-Al matrix, which are the same precipitate since the β″ precipitate is needle-like. In Fig. 7, the number density of the β″ precipitate increases slightly with the increase in Si content. Figure 8a shows the HRTEM image of the β″ precipitate embedded in the (001)_Al plane, and it clearly presents the unit cell of C-centred monoclinic structure with a = 1.52 nm and c = 0.67 nm, which verifies that the embedded precipitate is β″ [27, 28]. Figure 8b shows the corresponding FFT patterns of the rectangle area in Fig. 8a, and it also confirms that the embedded precipitate is β″. Figure 8c shows the HRTEM image of the β″ precipitate lying on the (001)_Al plane, and Fig. 8d shows the corresponding FFT patterns of the rectangle area in Fig. 8c, which verifies that the lying precipitate is β″, and the β″ precipitate is coherent with the α-Al matrix along the b-axis. The needle-like β″ precipitate provides peak strengthening effect [29, 30], which indicates that the T6 heat-treated Al–xSi–0.45Mg alloys are in the peak strengthening state.

Figure 7

Bright-field TEM micrographs showing the β″ precipitate in the T6 heat-treated Al–xSi–0.45Mg alloys: a 6.5 wt% Si, b 7.5 wt% Si, c 8.5 wt% Si and d 9.5 wt% Si

Figure 8

HRTEM micrographs taken along the <001> Al axis showing the β″ precipitate in the T6 heat-treated Al–xSi–0.45Mg alloys, a HRTEM image of embedded β″ precipitate, b FFT pattern of a, c HRTEM image of lying β″ precipitate and d FFT pattern of c

Mechanical properties after heat treatment

Figure 9a, b shows the tensile stress–strain curves and tensile properties of the Al5Ti1B-refined Al–xSi–0.45Mg alloys, after T6 heat treatment. Under the refinement of Al5Ti1B, with the increase in Si content from 6.5 to 7.5, 8.5 and 9.5 wt%, the yield strength (YS) increases nearly linear from 294 ± 2 to 299 ± 2, 304 ± 1 and 309 ± 2 MPa, and the tensile strength (UTS) increases from 336 ± 7 to 351 ± 4, 358 ± 3 and 363 ± 4 MPa, while the elongation (El) first increases slightly from 3.5 ± 0.8 to 4.5 ± 1.0%, then increases significantly to 7.8 ± 1.4%, after decreases to 5.5 ± 1.2%. Figure 9c, d shows the tensile stress–strain curves and tensile properties of the Al3Ti3B-refined Al–xSi–0.45Mg alloys, after T6 heat treatment. Under the refinement of Al3Ti3B, with the increase in Si content from 6.5 to 7.5, 8.5 and 9.5 wt%, the YS also increases nearly linear from 300 ± 1 to 305 ± 2, 312 ± 1 and 317 ± 2 MPa, and the UTS increases from 352 ± 3 to 360 ± 3, 367 ± 3 and 372 ± 3 MPa, while the elongation increases from 6.1 ± 1.1 to 8.5 ± 1.2, 11.8 ± 1.5 and 12.1 ± 1.6%. The Al3Ti3B-refined alloys have both higher strength and ductility than the Al5Ti1B-refined alloys. The YS and UTS of the alloys increase with increasing Si content. The ductility shows inverted ‘V’-shaped evolution with Si content and reaches the peak at 8.5 wt% Si when refined by Al5Ti1B, while the ductility increases with Si content when refined by Al3Ti3B.

Figure 9

a, c Tensile stress–strain curves and b, d tensile properties of the Al–xSi–0.45Mg (x = 6.5, 7.5, 8.5, 9.5) alloys refined by a, b Al5Ti1B and c, d Al3Ti3B after T6 heat treatment

Discussion

Si poisoning on microstructure under Al5Ti1B

Figure 10a shows the XRD pattern of the Al5Ti1B master alloy used for refinement; TiAl₃ and TiB₂ particles were found coexisting in the master alloy, which is consistent with the report that the particles introduced into the melt through the addition of Al5Ti1B are the soluble TiAl₃ and the insoluble TiB₂ particles [12]. Si in the melt reacted with Ti to form Ti–Si compounds, and the TiB₂ particles that act as heterogeneous nucleation sites for primary α-Al phase could be poisoned by Si by coating the surfaces with Ti–Si compounds [16, 17]. The detail poisoning mechanism of Si on TiB₂ particles has been verified by experiment recently with Al–Si–Ti particles observed on the prism face of TiB₂ [18]. The formation of Ti–Si compounds also consumes the Ti dissolved in the melt for grain growth restriction, and the solute Ti was reported hardly offering any grain growth restriction effect in Al–Si alloys with a Si content up to 7 wt% [15]. With the increase in Si content from 6.5 to 9.5 wt%, the poisoning effect of Si on the TiB₂ particles increases, and the heterogeneous nucleation of primary α-Al phase on TiB₂ particles becomes more difficult, which results in the continuous significant coarsening of the primary α-Al phase in the Al–xSi–0.45Mg alloys refined by Al5Ti1B. The increase in grain size coarsening rate in Al5Ti1B-refined Al–xSi–0.45Mg alloys from 7.5 wt% Si might be attributed to the loss of grain growth restriction.

Figure 10

X-ray diffraction patterns of a Al5Ti1B and b Al3Ti3B master alloys used for grain refinement

Si poisoning and promotion on microstructure under Al3Ti3B

Figure 10b shows the XRD pattern of the Al3Ti3B master alloy used for refinement; TiB₂ and AlB₂ particles were found coexisting in the Al3Ti3B master alloy. It was reported that AlB₂ alone without Si cannot effectively refine α-Al, while AlB₂ with the presence of Si could refine α-Al efficiently, indicating that the solute Si may interfere with AlB₂ to enhance its heterogeneous nucleating potential [20, 21]. It was speculated that the formation of unstable SiB₆ layer reduced the crystallographic mismatch between AlB₂ and Al, which enhanced the heterogeneous nucleating potency of AlB₂ for primary α-Al phase [22]. The formation of SiB₆ layer is still not verified by experiments, but the promotion of heterogeneous nucleation potency of AlB₂ by Si is the fact. There are two opposite effects of Si on the heterogeneous nucleation potency of TiB₂ and AlB₂ particles. With the increase in Si, the Si poisoning of the heterogeneous nucleation on TiB₂ particles increases, while the Si promotion of the heterogeneous nucleation on AlB₂ particles increases. For the Al3Ti3B-refined Al–xSi–0.45Mg alloys, the poisoning of Si on TiB₂ is not significant at 6.5 wt% Si, resulting in the fine primary α-Al grain size of 215 ± 30 μm; with the increase in Si content to 7.5 wt%, the poisoning of Si on TiB₂ is a little more significant than the promotion of Si on AlB₂, which leads to the slight increase in primary α-Al grain size to 265 ± 35 μm; with the further increase in Si content to 8.5 wt%, there is a balance between the poisoning of Si on TiB₂ and the promotion of Si on AlB₂, which maintains the primary α-Al grain size; with the increase in Si content to 9.5 wt%, the poisoning of Si on TiB₂ is again a little more superior than the promotion of Si on AlB₂, which causes the slight increase in primary α-Al grain size to 315 ± 25 μm. The sole nucleation site of TiB₂ suffers from enhancing Si poisoning with increasing Si under the refinement of Al5Ti1B, while the AlB₂ nucleation site benefits from continuous Si promotion with increasing Si besides the Si poisoning of TiB₂ nucleation site under the refinement of Al3Ti3B, which results in the significant finer grain size of primary α-Al and the slight coarsening of the primary α-Al with increasing Si in the Al3Ti3B-refined alloys, comparing with the Al5Ti1B-refined alloys.

Multiple effects of Si on mechanical properties

Effects on yield strength

The strengthening mechanisms in aluminium alloys generally include secondary phase strengthening, solution strengthening, precipitate strengthening, grain size strengthening and strain strengthening. For the T6 heat-treated Al–xSi–0.45Mg alloys, the yield strength is mainly controlled by the secondary phase strengthening of Si phase, the precipitate strengthening of β″ precipitation phase and the grain size strengthening of primary α-Al phase. In Figs. 5 and 6, the secondary eutectic Si phase in the as-cast Al–xSi–0.45Mg alloys was fully spheroidised after T6 heat treatment. So the volume fraction of the secondary Si phase in the as-cast alloys is the same as the volume fraction of the spheroidised Si phase in the T6 heat-treated alloys, which can be used for the evaluation of the secondary phase strengthening in the T6 heat-treated alloys. In Fig. 6, the β-Mg₂Si intermetallic phase was fully dissolved into the α-Al matrix after the solution treatment. In Figs. 7 and 8, the dissolved β-Mg₂Si phase precipitates in the form of β″ precipitate in the α-Al matrix for the precipitation strengthening of the alloys after the ageing treatment. Thus, the ratio between the volume fraction of Mg₂Si phase and α-Al phase in the as-cast alloys is the same as the ratio between the volume fraction of β″ precipitate and primary α-Al phase in the T6 heat-treated alloys, which can be used for the evaluation of the precipitate strengthening in the T6 heat-treated alloys. Figure 11 shows the evolution of the volume fraction of secondary Si phase and the ratio between volume fraction of β-Mg₂Si phase and α-Al phase with Si in the as-cast Al–xSi–0.45Mg alloys, which were calculated by the multicomponent phase diagram calculation software Pandat. With the increase in Si content, the volume fraction of secondary Si phase increases linearly, and the ratio between the volume fraction of β-Mg₂Si phase and α-Al phase increases nearly linearly. Thus, the secondary phase strengthening of spheroidised Si phase and the precipitate strengthening of β″ precipitation phase increase with increasing Si content in the T6 heat-treated alloys.

Figure 11

Volume fraction of eutectic Si phase and ratio between volume fraction of Mg₂Si phase and primary α-Al phase in as-cast Al–xSi–0.45Mg (x = 6.5, 7.5, 8.5, 9.5) alloys calculated by Pandat software

In Figs. 2, 3 and 4, it can be expected that the grain size of primary α-Al in the Al5Ti1B-refined alloys increases significantly with increasing Si, and the grain size of primary α-Al in the Al3Ti3B-refined alloys increases slightly with increasing Si, after T6 heat treatment, since T6 heat treatment hardly has any effect on the grain size. According to the Hall–Petch relation, the grain size strengthening decreases with increasing grain size. The decrease in the grain size strengthening with increasing Si in the T6 heat-treated alloys refined by Al3Ti3B is slighter than the alloys refined by Al5Ti1B. Under the refinement of both Al5Ti1B and Al3Ti3B, with the increase in Si, the increase in the secondary phase strengthening of spheroidised Si phase and the precipitate strengthening of β″ precipitation phase is superior to the decrease in grain size strengthening, which results in the increase in the yield strength with increasing Si, as shown in Fig. 9.

Effects on tensile strength and ductility

The tensile strength and ductility of the T6 heat-treated cast Al–Si–Mg alloys without porosity or other casting defects depend on the scale of the dendritic structure and the size and shape of the Si particles [31, 32]. The tensile strength and ductility of the T6 heat-treated cast Al–Si–Mg alloys with defects present are determined by the size and area fraction of defects on the fracture surface, rather than the bulk volume percentage of defects, and the tensile strength and ductility decrease monotonically with an increase in the area fraction of defects on the fracture surface [33, 34].

Figure 12a shows the evolution of the spiral flow length of the Al–xSi–0.45Mg alloys versus Si content under the standard fluidity tests. The spiral flow length increases with the increase in Si content, which indicates that the fluidity of the Al–xSi–0.45Mg alloys increases with increasing Si. Figure 12b shows the volume percentage of porosity in the alloys refined by Al5Ti1B and Al3Ti3B. Under the refinement of Al5Ti1B, with the increase in Si content from 6.5 to 7.5, 8.5 and 9.5 wt%, the porosity percentage first decreases slightly from 0.22 ± 0.02 to 0.18 ± 0.02%, then decreases significantly to 0.11 ± 0.01%, after increases to 0.15 ± 0.01%. Under the refinement of Al3Ti3B, with the increase in Si content from 6.5 to 7.5, 8.5 and 9.5 wt%, the volume percentage of porosity first decreases slightly from 0.13 ± 0.01 to 0.1 ± 0.01%, then decreases significantly to 0.024 ± 0.008%, after decreases slightly to 0.017 ± 0.006%.

Figure 12

a Spiral flow length of Al–xSi–0.45Mg alloys under the standard fluidity tests and b porosity percentage in the Al–xSi–0.45Mg alloys refined by Al5Ti1B and Al3Ti3B

Figure 13a–d presents the SEM images showing the fracture morphology in the Al5Ti1B-refined Al–xSi–0.45Mg alloys with 6.5 wt% Si, 7.5 wt% Si, 8.5 wt% Si and 9.5 wt% Si, respectively, after T6 heat treatment. Porosity defect was found on the fracture surface of the Al5Ti1B-refined alloys, and the insert in each figure shows the porosity morphology with higher magnification. With the increase in Si content from 6.5 to 7.5, 8.5 and 9.5 wt%, the size and area fraction of porosity on the fracture surface first decrease, then reach the minimum at 8.5 wt% Si, after increase, which is consistent with the evolution of the porosity percentage with Si content shown in Fig. 12b. From the insert in each figure, the grain size in the Al5Ti1B-refined alloys increases with increasing Si, which is consistent with the microstructure and statistical results of grain size shown in Figs. 2 and 4.

Figure 13

SEM images showing fracture morphology in the T6 heat-treated Al–xSi–0.45Mg alloys refined by Al5Ti1B a 6.5 wt% Si, b 7.5 wt% Si, c 8.5 wt% Si and d 9.5 wt% Si

Figure 14a–d presents the SEM images showing the fracture morphology in the Al3Ti3B-refined Al–xSi–0.45Mg alloys with 6.5 wt% Si, 7.5 wt% Si, 8.5 wt% Si and 9.5 wt% Si, separately, after T6 heat treatment. Porosity defect was found on the fracture surface of Al–6.5Si–0.45Mg and Al–7.5Si–0.45Mg alloys, and the inserts in Fig. 14a, b show the porosity morphology with higher magnification. With the increase in Si content from 6.5 to 7.5 wt%, the size and area fraction of porosity on the fracture surface decrease. With the further increase in Si content to 8.5 and 9.5 wt%, the porosity defect disappears from the fracture surface. The inserts in Fig. 14c, d show the enlarged fracture morphology, and the fracture comprises uniform distributed Al dimples and cracked Si, which is very similar to the reported Al3Ti3B-refined Al9SiMg alloy [10]. The evolution of porosity on the fracture surface of the Al3Ti3B-refined alloys with Si is consistent with the evolution of the porosity percentage shown in Fig. 12b.

Figure 14

SEM images showing fracture morphology in the T6 heat-treated Al–xSi–0.45Mg alloys refined by Al3Ti3B a 6.5 wt% Si, b 7.5 wt% Si, c 8.5 wt% Si and d 9.5 wt% Si

For the hypoeutectic Al–Si cast alloys, the porosity defect is mainly dependent on the solidification interval of the alloy, the fluidity of the liquid alloy and the grain size. Smaller solidification interval will result in lower tendency of porosity formation. Higher fluidity and smaller grain size will make the compensation of shrinkage easier and decrease the tendency of porosity formation. The solidification interval of the hypoeutectic Al–xSi–0.45Mg alloys decreases with increasing Si, which indicates that the tendency of porosity formation decreases with increasing Si from the viewpoint of solidification interval. In Fig. 12a, the fluidity of the liquid Al–xSi–0.45Mg alloys increases with increasing Si, indicating that the tendency of porosity formation also decreases with increasing Si from the viewpoint of fluidity. According to Figs. 2 and 4, the grain size in the Al5Ti1B-refined alloys increases significantly with increasing Si due to the enhancing Si poisoning of TiB₂ nucleation site, which indicates that the tendency of porosity formation in the Al5Ti1B-refined alloys increases with increasing Si from the viewpoint of grain size. In Figs. 3 and 4, the grain size in the Al3Ti3B-refined alloys increases slightly with increasing Si due to the enhancing Si promotion of AlB₂ nucleation site besides Si poisoning, which indicates that the tendency of porosity formation in the Al3Ti3B-refined alloys increases slightly with increasing Si from the viewpoint of grain size.

For the Al5Ti1B-refined Al–xSi–0.45Mg alloys, with the increase in Si content from 6.5 to 8.5 wt%, the decrease in porosity formation by decreasing solidification interval and increasing fluidity is superior to the increase in porosity formation by increasing grain size, resulting in the decrease in size and area fraction of porosity on the fracture surface and the consequent increase in tensile strength and ductility till 8.5 wt% Si; after the decrease in porosity formation by decreasing solidification interval and increasing fluidity is inferior to the increase in porosity formation by increasing grain size, resulting in the increase in size and area fraction of porosity on the fracture surface and the consequent decrease in ductility at 9.5 wt% Si. For the Al3Ti3B-refined Al–xSi–0.45Mg alloys, with the increase in Si content from 6.5 to 9.5 wt%, the decrease in porosity formation by decreasing solidification interval and increasing fluidity is superior to the increase in porosity formation by slightly increasing grain size, which leads to the consecutive decrease in size and area fraction of porosity on the fracture surface and the consequent continuous increase in tensile strength and ductility.

Conclusions

The effects of Si poisoning and promotion on the microstructure and mechanical properties of hypoeutectic Al–xSi–0.45Mg (x = 6.5, 7.5, 8.5, 9.5) cast alloys were investigated. The main conclusions are summarised as follows:

• 1. Al3Ti3B is superior to Al5Ti1B for the grain refinement of the Al–xSi–0.45Mg (x = 6.5, 7.5, 8.5, 9.5) alloys. With the increase in Si, Si poisoning on TiB₂ results in the obvious coarsening of primary α-Al in Al5Ti1B-refined alloys from 350 ± 40 to 400 ± 50, 475 ± 50 and 560 ± 80 μm, and the competition between Si promotion on AlB₂ and Si poisoning on TiB₂ leads to the slight coarsening of primary α-Al in Al3Ti3B-refined alloys from 215 ± 30 to 265 ± 35, 265 ± 30 and 315 ± 25 μm.

• 2. The strength and ductility of Al3Ti3B-refined alloys are superior to that of the Al5Ti1B-refined alloys, after T6 heat treatment. With increasing Si, the yield strength (YS) of Al5Ti1B-refined alloys increases from 294 ± 2 to 299 ± 2, 304 ± 1 and 309 ± 2 MPa, and the elongation first increases from 3.5 ± 0.8 to 4.5 ± 1.0 and 7.8 ± 1.4%, after decreases to 5.5 ± 1.2%, while the YS of the Al3Ti3B-refined alloys increases from 300 ± 1 to 305 ± 2, 312 ± 1 and 317 ± 2 MPa, and the elongation increases from 6.1 ± 1.1 to 8.5 ± 1.2, 11.8 ± 1.5 and 12.1 ± 1.6%.

• 3. The increase in the secondary phase and precipitation strengthening is superior to the decrease in grain size strengthening, which results in the increase in strength with increasing Si. With the increase in Si, the decrease in porosity formation by decreasing solidification interval and increasing fluidity is superior to the increase in porosity formation by slightly coarsening grain size, which leads to the continuous increase in ductility in the Al3Ti3B-refined alloys, while the competition between porosity decreasing and increasing factors leads to the inverted ‘V’-shaped evolution of ductility in the Al5Ti1B-refined alloys.