On the Effectiveness of Rotary Degassing of Recycled Al-Si Alloy Melts: The Effect on Melt Quality and Energy Consumption for Melt Preparation

Abstract

The effectiveness of rotary degassing on the defect formation and mechanical properties of the final casting of aluminium alloy EN AC 46000 was investigated, along with its impact on the energy consumption in the casting furnace. In the melt preparation prior to casting, the molten metal is usually transported from the melting furnace to the casting furnace with rotary degassing as a cleaning procedure. Under the conditions of this specific study, negligible degradation was observed in the mechanical properties of the final cast component in an aluminium EN AC 46000 alloy after removing the rotary degassing step in the process. Furthermore, removing the rotary degassing step led to a reduced temperature drop in the melt, thus minimizing the need for reheating (energy consumption) by up to 75% in the casting furnace. The reduced energy consumption was up to 124,000 kWh in yearly production in a 1500 kg casting furnace. The environmental impact showed a ~1500 kg reduction in CO₂ for one 1500 kg electrical casting furnace in a year.

1. Introduction

The usage of recycled aluminium alloys has increased because of global demands for higher resource efficiency and lower energy consumption. Consumers have shown that aluminium is a choice for a sustainable future since the recycling of aluminium only consumes 5% of the energy compared to primary aluminium production. Furthermore, energy usage and CO₂ footprint due to legislation and regulations (by means of emission standards since 1992) have served as the driving force to reduce emissions [1].

Nevertheless, the drive for reduced energy consumption per kg of aluminium is less highlighted in high-pressure die casting (HPDC) foundries. Most of the energy consumption in foundries is during production. Such energy consumption levels can be adjusted with different types of furnaces, capacity, and fuel [2]. The CO₂ footprint depends on several parameters, such as component weight reduction, electricity generation usage for material production, and melting/casting in the foundry [3].

Selecting lighter materials with high mechanical properties is essential for fuel/energy consumption in many industrial sectors, such as the transportation industry [4,5]. Moreover, recycled aluminium alloys are preferred for a circular society [6,7,8,9,10]. Al-Si-(Cu) cast alloys are a valid candidate to fulfil these requirements because of their high strength-to-weight ratio, cost-efficient manufacturing, good castability, and they are suitable candidates for recycling and circularity [6,7,9,11,12]. This is one of the reasons why secondary alloys with moderate mechanical property requirements, such as the EN-AC-46000 alloy, are commonly used in high-pressure die casting applications. However, significant variation in mechanical properties has been reported in these alloys, which has been attributed to (i) low melt quality, i.e., damaged liquid metal, (ii) un-optimized melting/casting process parameters, and (iii) improper post-treatment [13,14,15,16,17,18].

In many foundries, melting takes place in a tower furnace with a central location, and liquid metal is transported in ladles to die casting machines with their holding furnaces. The typical handling of the liquid metal in a die casting foundry is depicted schematically in Figure 1. Usually, no liquid treatment is given to the metal in the tower furnace, where rejected castings and filling and gating systems from previously produced castings are loaded along with virgin secondary ingots. Therefore, oxide surfaces on rejected castings, on gating systems, and on ingots are entrained into the melt. These entrained oxide films reduce the quality of the liquid metal [19] and reduce the casting performance by causing premature fractures [20].

To be transported to the casting furnace, liquid metal is allowed to drop from a height usually exceeding one meter into the transfer ladle. Consequently, oxide surfaces on the liquid metal, as well as air, are entrained into the bulk liquid in the transport ladle. Subsequently, a rotary degasser is immersed into the melt to pump either Argon or Nitrogen bubbles into the melt with a rotating impeller. Usually, a flux addition accompanies rotary degassing to limit the damage of entrained oxides. These bubbles collect dissolved hydrogen, as well as oxides and inclusions from the liquid metal, if the impeller rotates at an angular velocity that does not lead to further entrainment damage. Hence, optimization of the parameters for rotary degassing, such as gas flow, rotation speed, time, and melt temperature, is essential to improve melt quality [21,22]. It should be noted that rotary degassing may lead to lower metal quality if not conducted in the correct way [21,23]. Whether further entrainment occurs or not, rotary degassing usually leads to reduced hydrogen content in the melt, which has been accepted as a measure of melt quality improvement as a standard practice [23]. However, this may create a difficult scenario, in which melt quality is degraded due to surface entrainment during the rotary degassing process. In such cases, the reduced pressure test (RPT) that is commonly used to quantify melt quality gives the false result of improved quality due to reduced hydrogen content; in other words, the amount of oxide films in the liquid might be high, but they do not become visible in the RPT results because of reduced hydrogen content that minimized the porosities in the RPT [24]. After degassing, liquid metal is transferred into the holding furnace. This second melt transfer is also quite damaging to the metal due to the entrapped air during pouring from one furnace to the other. Furthermore, the melt in the holding furnace is in contact with the environment and the refractories and crucibles that increase the hydrogen content in the holding furnace. Moreover, evaluating the quality of the molten metal has been investigated for decades with several pieces of equipment [25]. However, work by Riestra et al. [26] showed that mechanical testing shows the most reliable results in correlation with melt quality.

In addition to the technical considerations for improving the melting and transportation of the molten metal, energy consumption needs to be considered for the sustainable foundry industry. Favi et al. [27] performed analytical cost estimation models in the HPDC process, taking cost, process parameters, product features, and accessory and setting up costs in their analysis. The energy cost was based on thermodynamical formulas depending on the furnace and the amount of internal remelting. However, the temperature of the molten metal during transfer was constant, and variations were not taken into consideration. Bonollo et al. [28] highlighted the importance of maintaining the temperature at a desired level in the casting furnace to minimize the number of defects in the final component. The energy-intensive foundry industry needs to work with energy savings and the environmental impact of the energy source, for example, fossil-based or hydro-based [29]. The CO₂ footprint for hydroelectricity is 20% that of fossil-based electricity. Depending on where in the world the electricity is produced will significantly impact CO₂ emissions [30]. In Sweden, one-kilowatt hour (kWh) generates 12 g CO₂ compared to the average in Europe, with 260 g CO₂, China 541 g CO_2, and 728 g CO₂ in Poland in 2020 [31].

The current paper aims to investigate the impact of rotary degassing of a secondary AlSi₉Cu₃(Fe) cast alloy (EN-AC-46000) on the energy consumption in an HPDC foundry, considering the mechanical properties of the final product as a criterion. The energy consumption of the material is investigated during melt handling before casting, while the mechanical performance is evaluated in the final component. Weibull analysis was used with a novel approach to present how the extent of damage evolves in the two process routes. The result of the mechanical properties shows no significant difference between the different routes.

2. Materials and Methods

The chemical composition of the secondary AlSi₉Cu₃(Fe) cast alloy (EN-AC-46000) was determined using a Spectromaxx CCD LMXM3 optical emission spectrometer, and the composition is presented in

Table 1. Chemical composition (wt.%) of the investigated alloy.

Alloy	Si	Mg	Cu	Fe	Mn	Zn	Al
EN-AC-46000	9.3	0.18	2.80	0.73	0.28	0.55	Balance

.

The melt was produced following the standard industrial practice in a tower furnace where recycled ingots, as well as rejects and filling systems of castings, are melted, following the sequence presented in Figure 1. Subsequently, flux (Coveral 2515) was added to the melt during the filling of the transport ladle to clean, and a degassing graphite impeller was immersed. As a common practice in the industry, for the rotary degassing, the rotor rotated at 450 rpm for 240 s, infusing 99.9% dry nitrogen. After the flux and degassing process, the slag on the melt surface was removed. Then, the transport ladle, filled with approximately 400 kg melt, was transported to where the 800-ton HPDC machine was located, and the melt was poured into the casting furnace for subsequent casting.

One set of melting/casting was conducted without the rotary degassing process to evaluate the effect of the degassing step. Samples to investigate the melt quality were taken from the casting furnace. Six RPT samples were taken to assess the liquid metal quality at each process route, three at a reduced pressure of 80 mbar, and three at atmospheric pressure. The density index was calculated based on the densities of the atmospheric and reduced pressure samples, measured using the Archimedes principle.

In order to understand the effect of degassing on the melt fluidity, five spiral fluidity tests were taken to further test the metal quality by manually filling the pouring cup above the spiral with a controlled volume and allowing the melt to flow in the spiral once the temperature reached 670 °C. Fluidity values were presented as the distance traversed within the spiral mould. Moreover, components from the production that passed the quality assurance protocols from the HPDC machine were cut, and five tensile specimens were excised. The parameters of the HPDC machine in production were kept constant to fulfil leak-free components. Furthermore, the temperature of the tower furnace and holding furnace followed the settings in the foundry. The microstructural characterization of the alloys was carried out by an Olympus DSX 1000 digital optical microscope. Tensile testing was carried out at room temperature, following the ASTM E8 standard, with a constant crosshead speed of 0.5 mm/min. Specimens were tested in Zwick/Roell Z100 tensile test equipment with a clip-on extensometer to measure the strain.

A Chauvin Arnoux PEL 103 power energy logger was used to record the energy consumption at the Sarlin 1500 kg ALU-hold furnace set to hold 685 °C temperature. The temperature of the melt in the transport ladle was measured with a Testo 735 logger using a type-N thermocouple.

3. Results and Discussion

3.1. Reduced Pressure Test

The density index (percent porosity) of the RPT specimens taken from the casting furnace, i.e., after one melt transfer following the rotary degassing treatment, was 6.45 ± 1.7 with rotary degassing and 6.96 ± 0.83 without rotary degassing. These results show marginal differences in the casting furnace with the two routes. The difference in density index reported in the literature [23] is commonly tested before pouring into the casting furnace. In contrast, samples in this study are taken in the production line with an additional melt transfer after the rotary degassing step, potentially negating any positive effect due to additional damage to the melt.

Figure 2a shows the RPT sample cross-section in the transport ladle before rotary degassing. Figure 2b shows a sample with significantly reduced porosities linked to the reduced hydrogen content in the melt from the rotary degassing process. However, as shown in Figure 2c,d, no significant changes are observed with or without rotary degassing after transfer to the casting furnace. This indicates that the quality of the melt before the additional pouring from the transfer ladle to the casting furnace shows a limited effect on the RPT samples in the casting furnace.

Thus, the rotary degassing step, under the circumstances of this study, has only minimal influence on improving the density index because of additional melt transfer that takes place after the degassing step, which reduces the melt quality again, as shown in Figure 2c,d, which is in line with the result from Kimura et al. [32].

3.2. Fluidity

The fluidity test was taken in the casting furnace and the results follow the RPT result, with essentially no difference between the two routes, showing the stability of the fluidity test. The test length in the route with and without degassing is 27.6 ± 0.75 cm and 27.7 ± 0.80 cm, respectively. The fluidity is reported by Kwon et al. [33] to decrease with more oxide films in the melt, as expected in the route of the degassing. However, since the values are within the standard deviation, could these results indicate that the already damaged melt from the beginning of the melting process in the production is very difficult to clean through flux additions and the rotary degassing process? Furthermore, this could also show that the rotary degassing process only removes the hydrogen in the melt, as shown in Figure 2b. In contrast, the fluidity test result shows no significant change. This could be due to the limited reduction in the oxide film, not removing but breaking the old oxides, which is in line with results from Dispinar and Campbell [34].

3.3. Mechanical Properties

The results of the tensile tests are presented in Figure 3. The tensile coupons excised from high-pressure die castings show that the yield strength (σ_Y) of the alloy exposed to degassing and flux is in the same range as the alloy without degassing (both with values around 150 MPa). Interestingly, tensile strength (S_T) and elongation to failure (e_F) are slightly higher for tensile bars when the rotary degassing process and flux are eliminated. In fact, the degassing should reduce the hydrogen content with a lower amount of porosities in the final component with improved mechanical properties. However, this seems to have a marginal effect on the mechanical properties. However, there is a rather large scatter of results for UTS and elongation to failure for both process routes. This indicates again that the melt has probably been highly damaged from the melting furnace and that the rotary degassing process is ineffective under the circumstances and conditions of this study. However, these mechanical results are above the EN AC 46000 specifications, showing that the rotary degassing under these circumstances only affects the melt in the transport ladle visually with the RPT samples. At the same time, the mechanical properties of the component show that, with or without the degassing process, all results fulfil the EN AC 46,000 specifications.

To assess the extent of the damage to the liquid metal, the concept of ductility potential for light alloy castings has been developed before [35]. The maximum elongation, e_F(max), can be written as

$${\mathrm{e}}_{\mathrm{F}\left(\mathrm{m}\mathrm{a}\mathrm{x}\right)}={\mathsf{\beta }}_0-{\mathsf{\beta }}_1{\mathsf{\sigma }}_{\mathrm{Y}}$$

(1)

where β₀ and β₁ are best-fit constants, for Al-Si-Mg alloys, β₀ and β₁ are 36.0 and 0.064 MPa⁻¹, respectively. Therefore, the specimens with yield strengths of 124 and 150 MPa, as in this study, can be expected to have a maximum elongation of 28.1% and 26.4%, respectively. The low elongation to failure is due to extensive entrainment damage given to the metal throughout the production. Based on this concept, the structural quality index, Q_T, was developed [35];

$${\mathrm{Q}}_{\mathrm{T}}=\frac{{\mathrm{e}}_{\mathrm{F}}}{{\mathsf{\beta }}_0-{\mathsf{\beta }}_1{\mathsf{\sigma }}_{\mathrm{Y}}}$$

(2)

If Q_T is less than 0.25, the metal is considered highly damaged.

To assess the extent of damage in the two different routes in the process, tensile results were converted to Q_T. The average Q_T result values are in a range of 0.027–0.028. The two-parameter Weibull distribution was used to determine the statistical distribution of Q_T results. The cumulative probability function, P [36,37], of the Weibull distribution is written as

$$\mathrm{P}=1-\exp \left[-{\left(\frac{\sigma }{{\mathsf{\sigma }}_0}\right)}^{\mathrm{m}}\right]$$

(3)

where σ₀ is the scale parameter and m is the Weibull modulus. These two parameters were calculated using the linear regression method [38,39]. The goodness-of-fit test was conducted to determine whether data come from a Weibull distribution by following the method [40] that uses the coefficient of determination, R². The results of the Weibull analysis are presented in

Table 2. Results of the 2-parameter Weibull analysis of the quality index results.

	N	m	σ0	R2	Weibull
With Degassing	9	3.77	0.14	0.93	Yes
Without Degassing (low)	7	8.38	0.11	0.92	Yes
Without Degassing (high)	4	17.95	0.22	0.98	Yes

. In the two cases, data follow the Weibull distribution. However, it should be noted that a Weibull mixture [41,42], i.e., two Weibull distributions, was noticed in samples collected without degassing. These are indicated as low and high in Table 2. Therefore, these two distributions have no overlap and could be analysed separately.

To characterize the structural quality and the damage given to the liquid metal at every stage, the Weibull distributions of Q_T results were plotted by using the probability density function, f;

$$\mathrm{f}=\frac{\mathrm{m}}{{\mathsf{\sigma }}_0}{\left(\frac{\sigma -{\sigma }_t}{{\mathsf{\sigma }}_0}\right)}^{\mathrm{m}-1}\exp \left[-{\left(\frac{\mathsf{\sigma }-{\mathsf{\sigma }}_{\mathrm{t}}}{{\mathsf{\sigma }}_0}\right)}^{\mathrm{m}}\right]$$

(4)

Results are presented in Figure 4, where the degassing route in the casting furnace shows the wide distribution in the Q_T results compared with the material without the degassing process, showing two peaks and using the Weibull distributions of the structural quality index, meaning the rotary degassing process eliminates the Weibull mixture and produces only one distribution. However, it should be noted that the “high” distribution, indicating less damage is eliminated, indicated that the metal is damaged during rotary degassing. This is further demonstrated by the fact that the lower tail of the distribution for degassed specimens is lower than that of the “low” distribution of specimens produced without degassing. Applying the Weibull distributions of the structural quality index is a new way to show how the melt quality changes in production. The authors believe that the damage in the melt was so severe that the rotary degassing impeller could only break the old coarse oxides but did not remove them from the melt. This might require further experimental validation but is in line with results from Dispinar and Campbell [34] and Gyarmati et al. [23].

3.4. Energy Consumption

Since the mechanical properties show minor differences between the two process routes in the foundry, it is interesting to investigate how the energy consumption varies. The energy consumption in the casting furnace mainly depends on the temperature of the melt from the central tower furnace. There is a big difference in the temperature of the melt if it is exposed to rotary degassing treatment or not. The temperature was 680 ± 5 °C with degassing, compared to 715 ± 5 °C without degassing, measured with thermocouples in the transport ladle. The set value of the casting furnace in the process is 685 °C. The cast component produced in the HPDC machine has a cycle time of 1 min and 50 s, with a weight of around 4 kg. The furnace was filled with about 400 kg of melt every third hour in the production shown by the red arrows in Figure 5. The energy consumption depends on the temperature of the melt added to the casting furnace. Using the rotary degassed melt with a temperature of around 680 °C reduces the temperature of the casting furnace to about 670 ± 2 °C. The furnace starts to heat the melt to reach the set value, which takes more than three hours, measured with the PEL 103 power energy logger. This result shows that the casting furnace must continually heat the added melt until the next filling time. The solid line demonstrates this in Figure 5.

The temperature of the casting furnace after three hours is 683 ± 2 °C. The energy consumption during the heating is 18.8 ± 0.3 kWh. However, the route without degassing significantly reduces time from more than 3 h to 45 min for the casting furnace to reach the set temperature (see the dashed line in Figure 5). This shows the importance of the melt temperature, whereas even an addition of 400 kg melts with a temperature of 715 ± 5 °C reduces the temperature of the casting furnace to about 681 ± 2 °C. This difference reduces the energy consumption by approximately 75% (see the green arrow in Figure 5). The 1500 kg casting furnace is used 24 h per day, and with these energy measurements, a yearly production with the degassing route will use approximately 165,000 kWh. However, the route without degassing resulted in an average annual reduction of about 124,000 kWh. Furthermore, this result shows that controlling the temperature in the casting process could significantly benefit an energy-consuming industry. The melting station can be optimized with this knowledge to reduce energy consumption. However, it is essential to note that this study only focused on the effect of rotary degassing; another approach to understanding is slag removal and maintenance of the furnace, which need to be investigated further to have a more comprehensive view of other effects. This reduction of ~124,000 kWh in one 1500 kg casting furnace for a year lowered the CO₂ emissions from the foundry located in Sweden with 12 g CO₂ per kWh, resulting in a total reduction of 1500 kg CO₂ per year [31]. Furthermore, this can be converted to 1.7 g CO₂ per kilogram cast AlSi₉Cu₃(Fe) in Sweden. Moreover, this significant reduction in CO₂ from energy consumption will also reduce the cost of energy and nitrogen gas for the foundry.

4. Conclusions

The current study of an aluminium EN-AC-46000 alloy with requirements of elongation to failure below 1% clearly shows that the rotary degassing process is not improving the properties of the material since the molten metal was already damaged from the beginning of the melting process. If the rotary degassing is removed from the process under the circumstances of this study, the following conclusions can be made.

• The mechanical properties of the cast component show a slight trend of improvement rather than reduction.
• The novel way to use Weibull distribution shows how the melt quality changes in the process, whereas the degassing route eliminates the Weibull distribution mixture and produces only one distribution. This shows that the rotary degassing under these circumstances damages the melt.
• The melting and casting process had a 75% reduction in energy consumption in the investigated set-up, saving around 124,000 kWh in the yearly production of a 1500 kg casting furnace.
• The removal of the rotary degassing reduces the CO₂ emissions by 1500 kg from one casting furnace in Sweden per year, ~32,000 kg in Europe on average, and ~93,000 kg in Poland. Moreover, the settings and parameters in this study are equivalent to 1.7 g CO₂ per kilogram cast AlSi₉Cu₃(Fe) in Sweden.

This result shows that with minor adjustments in the production flow, a significant reduction in CO₂ emissions with no decrease in mechanical properties is achieved. Moreover, implementing this in the foundries should be beneficial since it reduces energy costs and shortens the transport time from the tower furnace to the casting furnace.