Ductility Potential and Quality Index for Aluminum Alloy Castings: An Update

Abstract

An update to the ductility potential of Al-Si and all other cast aluminum alloys is provided. Through analysis of extreme data in the literature, it is demonstrated that the highest elongation values in cast aluminum alloys are quite similar to those in wrought aluminum alloys. Through meticulous attention to all details of liquid metal quality throughout the entire production system, castings with exceptional quality can be produced.

1. Introduction

During a tensile test, there is competition within the specimen between further deformation and fracture [1], the outcome of which is determined by its defect content. If the energy needed by the largest defect to cause premature fracture exceeds that needed for further deformation, tensile testing continues and the specimen remains deforming, reaching higher levels of strain. Otherwise, metal fractures prematurely due to stress concentration near the largest defects. Therefore, the energy absorbed by the specimen during a tensile test is a function of the structural quality of the metal, i.e., the extent of damage given to the metal in liquid state. Liquid metal damage refers to entrainment of the surface oxide film into the bulk liquid during the entire production chain, including ingot production, melting, melt processing and mold filling, resulting in common casting problems, such as pores, hot tears and low ductility and fatigue life. The lower the liquid metal damage, the higher stresses and strains are reached during tensile testing, resulting in larger amounts of energy absorbed. Conversely, castings with a high level of liquid metal damage absorb much less energy during tensile testing, because of the higher number of defects that are large enough to win the competition to fracture early during tensile testing.

Strain energy density, alternatively referred to as toughness, is the area under the true stress–true strain curve, and represents the energy absorbed by the material per unit volume. It has been shown [2,3] that elongation (e_F) is an excellent estimate of strain energy density in A357 and A206 aluminum alloy specimens. Elongation is known to be quite sensitive to the presence of defects [4], whereas yield strength (σ_Y) is minimally affected. Therefore, the authors [5,6,7] plotted data from aerospace and premium quality castings with elongation as the y-axis and yield strength as the x-axis, as recommended by Staley [8] and Lloyd [9]. Castings with the higher elongation at the same yield strength are considered to have fewer defects, i.e., higher structural quality. The authors have reported [5,6,7,10,11] that the maximum points for the Al-7%Si-Mg alloy family, A201 and 206 alloys, had a linear trend. The authors developed the concept of the ductility potential, e_F(max), for aluminum alloy castings:

$$e_{F\left(\mathrm{m}\mathrm{a}\mathrm{x}\right)}={\beta }_0-{\beta }_1{\sigma }_Y$$

(1)

where β₀ (%) and β₁ (%/MPa) are constants for each alloy. The originally reported values of these constants for the three alloy families are provided in

Table 1. The coefficients of Equation (1) for Al-7%Si-Mg, A201, and A206 alloys as originally presented [11].

	β0 (%)	β1 (%/MPa)
A356-357	36.0	0.064
A201	34.5	0.047
A206	47.8	0.085

.

The authors provided a quality index, Q_T, as the ratio of actual elongation to the ductility potential:

$$Q_T={\displaystyle \frac{e_F}{e_{F\left(\mathrm{m}\mathrm{a}\mathrm{x}\right)}}}$$

(2)

This quality index is shown schematically in Figure 1. This index provides a metric based on the actual energy absorbed by the specimen as opposed to how much energy it would absorb if there were minimal or no defects. Hence, Q_T is a measure of the damage given to liquid aluminum. Since its introduction, Q_T has been used with increasing popularity in research and production settings.

When the ductility potentials for the three alloys were initially established, there were no data available above these lines. Since then, several datasets have emerged, especially for the Al-Si alloys. This present study is motivated by the need for an update in ductility potential and the interpretation of quality index values.

2. Ductility Potential Update

The update in the present study is given in two categories: Al-Si alloys and Al alloys without Si. The latter consists of Al-Cu, Al-Mg, and Al-Zn alloy families, although only Al-Cu alloys are used in aerospace and premium castings. Therefore, only Al-Cu data are analyzed in the present study. However, all three alloy families have counterparts as wrought aluminum alloys. Therefore, the previously analyzed data for A201 and A206 are first combined and then compared with typical elongation values for all wrought aluminum alloys, as documented by Kaufman [12]. The combination of the data from the two Al-Cu-Mg alloys is presented in Figure 2. Note that the two datasets yield a single ductility potential line. The values for β₀ and β₁ for the updated ductility potential line for all cast Al alloys (excluding Al-Si) are provided in

Table 2. Values of the two coefficients for the ductility potential lines.

Alloy Family	β0 (%)	β1 (%/MPa)
Al alloys (no Si)	40.5	0.060
Al-Si	36.0

. Note that the slope for all cast aluminum alloys, including Al-Si, is now proposed as 0.060 (%/MPa). The common slope between the two datasets emerged during the reanalysis of data for the present study.

Typical elongation values at their corresponding yield strengths of commercial wrought aluminum alloys, as listed by Kaufman [12], are shown in Figure 3. Also indicated is the revised ductility potential for all aluminum alloys, excluding Al-Si. The ductility potential line in Figure 2 exceeds most of the data for wrought aluminum alloys. Only several data points are above the ductility potential line. When the metal has very low to no liquid metal damage, the ductility of cast and wrought alloys are at the same level, as can be expected.

Turning our attention to Al-Si alloys, these alloys (with Si above 1.6 wt.%) are by far the most commonly used alloys in castings and have a microstructure that consists of a primary aluminum matrix and an Al-Si eutectic. Unmodified eutectic Si particles have been observed [13] in situ to fracture early in tensile plastic deformation, and their fractures propagate and coalesce into macroscopic failure. This premature fracture is attributed to their size and shape, and Si particles are considered to be weak and brittle. As a result, ductility in cast Al-Si alloys has been believed [14] to be controlled by the morphology and size of Si particles. In situ mechanical testing of Si particles [15,16] has shown that, while some particles reach the theoretical strength of 16 GPa [17,18], others have fractured at significantly lower stresses. This is consistent with other studies [19,20,21] in the literature, although the lowest fracture stress reported varies between studies. This reduction in strength can be attributed to defects inside the Si particles, suggested at times to be pinhole defects [22], but are almost certainly usually bifilms [23] on which they have been found [24,25,26,27] to nucleate heterogeneously. Hence, some Si particles exhibit poor mechanical performance due to these artifacts from liquid metal damage during the casting process [28].

Modified eutectic Si behaves somewhat differently, growing not on bifilms but forming by a coupled growth with aluminum, the Si taking a coral-like form. In this case the bifilms are still present, and act to initiate the final failure from their random locations in the eutectic [29]. However, of course, liquid damage is still exerting its major influence, albeit in somewhat different ways, in both modified and unmodified Al-Si alloys.

Seven datasets are included in the present update for the Al-Si alloys. These new data are presented in Figure 4, with the previously analyzed data shown in black squares. The details of these datasets, including alloy composition and the casting method, are given in

Table 3. Details of the datasets of cast Al-Si alloys included in this study.

Dataset	Alloy	Process	Ref.	λ2 (μm)	Additional Notes
1	A356	Sand casting		50 *	Commercial foundry, slowly cooled, heat treated to T4
2	C355	Investment casting	[30]	38	Heat treated to T6
3	Al-7Si-Mg	Low-pressure casting	[31]	20	Heat treated to T6
4	A356	Casting **	[32,33]	42	Heat treated to T6
5	A356	Sand casting with end chill	[34]	8–14	Heat treated to T6
6	Al-7Si	Permanent mold semi-solid casting	[35]	N/A	Heat treatment according to modified strain-induced melt activation (M-SIMA) process
7	A356	Thixocasting	[36]	N/A	Not heat treated

*: estimated. **: type of casting process not provided.

. Note that most data come from studies on Al-7%Si alloys, such as A356 alloy. In addition to sand and investment casting, two semi-solid casting processes, strain-induced melt activation (SIMA) and thixocasting, are included in this analysis. The dashed line in Figure 4 represents Q_T = 1.30. Therefore, an excellent combination of strength and ductility, beyond what was reported previously only in aerospace and premium quality castings, is achievable in these alloys, provided that liquid metal damage is minimized or even eliminated.

3. Interpretation of Quality Index Numbers

With the update provided above, a revised interpretation of quality index values is also recommended, as presented in Figure 5. The values for Q_T < 0.25 remain as High Damage, indicating very poor melt quality consisting of major old oxides and poor metal handling processes. Q_T values between 0.25 and 0.50 indicate Moderate Damage, the presence of some old oxides, and additional damage in metal processing, such as melt transfers and turbulent degassing practices. Q_T values between 0.50 and 0.70 indicate Low Damage. Melts are free from major oxides and damage from metal processing and mold filling is low. When Q_T is between 0.70 and 1.00, the level of damage given to liquid metal is very low. Only very careful processing and mold filling practices will enable foundrymen to achieve these Q_T values. The authors think that values that exceed 1.00 can only be categorized as Exceptional Quality. This new region was coined to keep the potential ductility line for Al-Si alloys (almost) intact since it has gained wide acceptance in industry since its inception. More discussion about this region is provided below.

4. Microstructural Effects

It is well established that secondary dendrite arm spacing, λ₂, is a function of local solidification time [37,38,39,40]. Moreover, elongation has been found [41,42] to decrease with increasing secondary arm spacing in cast aluminum alloys. That is why measuring λ₂ in aluminum castings has even been suggested as a new nondestructive test in aerospace castings [43].

In one of the original studies [6] in which ductility potentials were established, the authors provided λ₂ numbers for the tensile data that were on or around the ductility potential line. Those data came from castings in which λ₂ varied between 13 and 45 μm. This has led the authors to conclude that λ₂ mattered only when bifilms are present in the metal. With increased local solidification time, bifilms that are compact at the end of mold filling unfurl under negative pressures developed during solidification, hydrogen diffusion and/or nucleation, and the growth of intermetallics on them. In the absence of bifilms, longer local solidification times did not matter.

The secondary dendrite arm spacing values of the new casting data are included in Table 3. These values, along with those originally reported [6] by the authors are presented in Figure 6. There is no correlation between λ₂ and Q_T in the “Very Low Damage” and “Exceptional Quality” regions. The solidification time of aluminum castings does not affect ductility when the metal is clean.

Although Al-Si alloys are widely used in castings, their use is limited as wrought alloys (4xxx series). As a result, only scarce data are available for wrought Al-Si alloys. Moreover, ductility data for Al-Si alloys that have gone through manufacturing processes other than casting are not readily available in handbooks. To provide a comparison between cast and wrought alloys similar to the one given in Figure 3, a literature survey was conducted. Nineteen datasets from the same number of independent studies have been collected. The details are provided in Table 4. Most data come from Al-7%Si alloys such as A356 and A357, but data from Si contents as low as 1.5 and as high as 20% are also included. Moreover, data for two 4xxx alloys are also reported. A total of ten different manufacturing methods have been used in these studies, including equal channel angular processing (ECAP) and friction stir processing (FSP). The results are presented in Figure 7. It is noteworthy that a majority of the data falls into the “Exceptional Quality” region defined earlier, with some data even exceeding Q_T = 1.30. There are many datapoints that fall in the Very Low Damage region. These are mostly the data for 4xxx alloys. Only two datapoints, both for the extruded 4032, fall below the Very Low Damage region.

It would be obvious to any metallurgist that the microstructures obtained in these processes would be vastly different from each other, and definitely from the microstructures of cast alloys in Figure 4. Yet, the yield strength–ductility correlation overlap almost completely. Moreover, two datapoints, one with no eutectic (1.5%Si) and the other eutectic with primary Si particles (Al-20%Si) fall in the same scatter band. The fact that the same excellent combination of strength and ductility can be obtained through various processes, which produce vastly different microstructures, suggests that the true effect of microstructure is at best, limited. It is noteworthy that processes that produce severe deformation, including ECAP and FSP, are known to increase the ductility of Al-Si alloys. This increase in ductility can be attributed to the destruction of bifilms, and not necessarily due to any refinement of Si particle size [44]. Obviously, more research is needed to determine the true microstructure–property relationships in Al-Si alloys.

Based on the analysis above, it can be stated that Exceptional Quality is an extension of the Very Low Damage region. Here, microstructural parameters, such as secondary dendrite arm spacing, Si particle size, and grain size are no longer correlated with mechanical properties. Alloy composition and processing determine the yield strength, which is always coupled with excellent ductility.

Table 4. Details of the datasets of other Al-Si alloys included in this study.

Dataset	Alloy	Process	Ref.
8	Al-10Si-Mg	Electron beam melting	[45]
9	Al-6Si-Cu	Hot extrusion	[46]
10	Al-7Si	ECAP	[47]
11	Al-20Si	ECAP	[48]
12	A356	FSP	[49,50]
13	A356	Hot extrusion	[51]
14	A356	Hot extrusion	[52]
15	A356	ECAP, cryo-rolled	[53]
16	A356	Twin roll cast	[54]
17	Al-3%Si-Fe	Twin roll cast	[55]
18	A356	Spinning deformed	[56]
19	A357	Selective recrystallization	[57]
20	Al-7Si	ECAP	[58]
21	Al-7Si	FSP	[58]
22	Al-1.5Si	ECAP	[59]
23	4032	Hot extrusion	[60]
24	4032	Hot extrusion	[61]
25	4032	Hot extrusion	[62]
26	4043	Rolling	[63]

Table 4. Details of the datasets of other Al-Si alloys included in this study.

Dataset	Alloy	Process	Ref.
8	Al-10Si-Mg	Electron beam melting	[45]
9	Al-6Si-Cu	Hot extrusion	[46]
10	Al-7Si	ECAP	[47]
11	Al-20Si	ECAP	[48]
12	A356	FSP	[49,50]
13	A356	Hot extrusion	[51]
14	A356	Hot extrusion	[52]
15	A356	ECAP, cryo-rolled	[53]
16	A356	Twin roll cast	[54]
17	Al-3%Si-Fe	Twin roll cast	[55]
18	A356	Spinning deformed	[56]
19	A357	Selective recrystallization	[57]
20	Al-7Si	ECAP	[58]
21	Al-7Si	FSP	[58]
22	Al-1.5Si	ECAP	[59]
23	4032	Hot extrusion	[60]
24	4032	Hot extrusion	[61]
25	4032	Hot extrusion	[62]
26	4043	Rolling	[63]

5. Conclusions

Ductility potential lines for cast aluminum alloys have been updated. This update presents only two equations, one for Al-Si alloys and the other for all other cast aluminum alloys. Al-Si alloys can show exceptional ductility for any yield strength level independent of the Si content and processing route, indicating the weak effect of microstructure. The ductility potential of cast aluminum alloys is comparable to the maximum ductility levels obtained in wrought alloys. Through meticulous attention to liquid metal quality throughout the entire production system, it seems entirely possible to produce castings with exceptional ductility.