Fluidity of Aluminium Foundry Alloys for Thin Wall Castings: Designing an Operating Methodology

Abstract

Aluminium thin wall castings are gaining wide acceptance in the automotive industry because of their incomparable design flexibility and higher mechanical properties. For these thin wall castings, fluidity plays a vital role in determining the quality of the final product. The aim of this work is to provide a detailed insight into the development of a multi-channel testing methodology to evaluate the fluidity of aluminium foundry alloys for thin wall applications. AlSi10MnMg foundry alloy has been used to conduct a series of experiments with the aim of designing operative protocols that achieve higher repeatability of the results. The fluidity of the investigated alloy was observed in channels of various cross-sections at three different pouring temperatures, i.e., 680, 710, and 740 °C. The obtained results show that experiments conducted following closely the designed operative protocols, result in achieving higher repeatability. It was also observed that by increasing the pouring temperature, the fluidity and repeatability of the alloy increased greatly. The 3D transient simulations were conducted by means of Altair^® Inspire™ Cast 2021.2 software to study the molten metal behaviour, i.e., solidification temperature and time at the end of each strip for the studied pouring temperatures. The results further reveal that the design methodology, if executed with intrinsic accuracy and precision, will provide a reliable pathway to determine the fluidity of aluminium alloys for various industrial applications.

1. Introduction

Thin wall castings are widely used in the aerospace and automotive industries because of their superior characteristics, i.e., complex design flexibility, weight reduction, cost efficiency, effective heat dissipation and high production rate. These castings significantly play a vital role in enhancing fuel efficiency and reducing emissions of vehicles [1,2,3]. Aluminium and its alloys in the automotive industry are considered an optimal choice for a range of applications, including thin wall products, because of their low density, high specific stiffness, high strength/weight ratio and excellent castability [4,5,6,7,8]. In particular, Al-Si-based alloys are widely used in the die-casting industry along with the addition of other alloying elements, i.e., Mg, Cu, Mn, Ti, etc., to enhance the overall casting performance, including high abrasion and corrosion resistance [9,10,11,12]. However, thin wall aluminium castings face several challenges due to the rapid cooling of thin sections, which could affect the molten metal flow during mould filling. Hence comprehensive understanding and thorough knowledge of molten metal flow within the mould cavity is essential for producing sound and defect-free components. In casting processes, molten metal continuously losses its temperature while filling the mould, which directly affects the metal flow behaviour and its solidification mechanism. Hence, molten metal flow and its solidification process are considered critical factors which ultimately define the overall quality and integrity of the final product. In the casting process, the filling of molten metal in the mould cavity is considered a critical step as metal loses temperature while filling. The behaviour of molten metal and its solidification mechanism plays a vital role in determining the quality and integrity of the final cast product. The poor filling leads to the formation of casting defects, which not only results in the rejection of casted components but also leads to an increase in manufacturing cost. Considering its importance, many researchers have investigated these phenomena, which led to the development of a quantitative concept for characterizing metallic material in the liquid state, called “fluidity”.

In the foundry process, fluidity is defined as the empirical measurement of the maximum length covered (in millimetres or centimetres) by the molten metal or alloy in a specific channel of constant cross-section area before it fully solidifies [13,14]. However, it cannot to regarded as a simple physical property, as it is a complex technological attribute. In physics, fluidity is defined as the reciprocal of viscosity; however, this definition cannot be directly applicable in the foundry context as its measure depends upon multiple factors [14]. It is widely acknowledged that higher fluidity in metal or alloys allows the production of more complex and intrinsic thin wall castings. Despite its critical importance, many aspects of this property are still unexplored. Therefore, a comprehensive understanding of all factors influencing the overall fluidity of metals and alloys is essential. The major factors influencing the fluidity of metals and alloys can be divided into two main categories: (a) melt-based factors, which include melt composition, viscosity, metallostatic pressure, melt superheat, surface tension, specific weight, melt cleanliness, solidification mechanism and (b) mould based factors including mould material, mould temperature, mould permeability, mould coating and mould surface tension. The fluidity measurement is dependent upon the optimization of these factors [15,16,17,18].

In foundry, various testing methods have been employed by engineers and researchers to study and measure the fluidity of molten metals. Among those, two are widely recognized: the spiral fluidity testing [19] and the vacuum fluidity testing [20]. Both testing methods work on a common principle, where molten metal flows into a channel with a constant cross-section area. In comparison to vacuum testing, the spiral testing technique gained more acceptance because of its mould compactness and less sensitivity of the procedure towards levelling error [21,22,23]. Although widely used in the foundries, this testing technique is often criticized for its lack of proper correlation with the real casting scenario, as the fluidity length measured through this method applies only to channel sections with a constant cross-section area [14]. Consequently, the single stream spiral testing procedure raises questions regarding the molten metal behaviour in real casting situations, where the mould geometry experiences multiple metal streams through various cross-section thickness branches and merges quite often. Moreover, as there are no standardized protocols, there is a large variation in fluidity results obtained at the same operating conditions, i.e., poor repeatability. This poor repeatability raises questions about the reliability of this testing method and consequently, reliable fluidity data for aluminium and its alloys is hard to find in the existing literature.

To provide a better understanding of the fluidity of metal and alloys for thin wall applications, a multi-channel testing approach was presented, where multiple strips of similar length but varying cross-sectional thickness are filled simultaneously from universal runner bars and flow length across strips with different cross-section areas can be measured simultaneously. Shin et al. [24,25] investigated the behaviour of molten metal through multichannel network fluidity testing to determine the castability of different aluminium alloys. These investigations highlight the importance of using multichannel moulds and the role that local thickness gradient plays in shaping the flow behaviour of molten metal. Campbell, in his work, summarizes that determining the fluidity as a function of channel cross-section thickness gives an essential understanding for establishing the effective surface tension for mould filling [26]. In addition to proficient fluidity evaluation for thin sections, the relative repeatability of this method is much greater as well. The fluidity investigation of Al-Mg-Si alloys was conducted by Sabatino et al. [27] through both spiral and strip testing methods. The results show that 11% of relative repeatability was obtained through strip testing, while only 5% was achieved through spiral testing. Adefuye et al. [28] studied the Al-Si alloys through sand mould multi-channel strip testing methods. Their results show that simple adjustments in design can generate results with errors less than 10% in strip thicknesses ranging from 4 mm to 1 mm. These findings result in raising the interest of the researchers in multi-channel fluidity testing methodology [29].

Although experimental testing is key to determining fluidity, each trial can be time-consuming and costly. Therefore, numerical simulation may be essential because of its efficiency and versatility. It helps in eliminating the need for extensive material usage and continuous physical testing, significantly reducing the experimental cost and saving time. Futáš et al. [30,31] have found no significant statistical difference between experimentation and simulation results. Numerical and experimental validation of AlSi12CuNiMg alloys was conducted by Khandelwal and Gautam through multi-spiral channels of varying thickness [32]. Their simulation results show very good agreement (91%) with the fluidity length obtained in their experimental trials. Žbontar et al. [33] also found a good correlation between numerical and experimental work for the fluidity of cast iron through multi-spiral channels with various thicknesses. In the literature, the molten metal flow behaviour of various alloys with multichannel of various cross-section thicknesses has been investigated; however, it is rare to find tests of AlSi10MnMg foundry alloy in multi-strip channels with various thicknesses.

Extensive work has been performed to study the effects of processing conditions and various alloying elements addition on the fluidity of alloys [22,34,35], while less attention has been paid to their operating methodology. The aim of this document is to present an operating methodology for evaluating the fluidity of aluminium alloys in thin wall castings in a robust and reliable manner, providing valuable insight for foundry applications. The adopted methodology is based on multi-channel testing, with the primary focus on establishing the operating protocols to enhance the repeatability of fluidity evaluation for aluminium and its alloys, which was successfully achieved because of good agreement between numerical simulation and experimental testing.

2. Experimental

2.1. Material

AlSiMnMg alloys are by far considered the most widely used to produce High Pressure Die Cast (HPDC) automobile structural components such as the shock tower, front pillars, etc. These alloys contain high silicon content that is highly suitable for casting complex shapes due to their enhanced fluidity, primarily attributed to the high volume of the eutectic Al-Si phase and silicon’s high enthalpy of fusion (1810 kJ/kg compared with 395 kJ/kg for aluminium). This higher enthalpy extends the duration in which the molten alloy remains fluid, allowing it to flow more effectively through the mould before solidifying [22,36]. In the present study, the fluidity of AlSi10MnMg alloy was investigated through permanent mould casting. The chemical composition of the investigated alloy was obtained on a sample sectioned from ingot and analyzed using SPECTROMAx LMX08 s (Ametek, Weiterstadt, Germany). Five readings were taken, and the average chemical composition is reported in

Table 1. Chemical composition (wt%) of AlSi10MnMg alloy.

Si	Fe	Cu	Mn	Mg	Cr	Ni	Ti	Sr	V	Al
11.480	0.170	0.018	0.657	0.143	0.005	0.003	0.056	0.02	0.010	Bal.

.

2.2. Fluidity Test

Multi-channel testing was used to measure the fluidity of the given alloy at different pouring temperatures. Figure 1 shows the mould geometry for the cope (a) and drag (b) parts of the experimental die made of H-13 steel. Since thin wall castings experience a high cooling rate, any variation in mould surface quality significantly affects the heat transfer by altering the temperature distribution and cooling rate during the casting process. This variation directly impacts the flow and solidification mechanism of molten metals [37,38]. Hence, both cope and drag were obtained with smooth surface finish, i.e., low surface roughness. The cope consists of six channels of different cross-sections 1, 3, 5, 7, 9, and 11 mm with constant length of 228 mm and width of 20 mm. The pouring basin, also made of H-13 steel, was placed over the cope. The cope and drag surfaces were fully covered with boron nitride aerosol lubricant, which provides high-releasing properties, preventing the melt from adhering to the mould surfaces and, as a result, increasing the efficiency of the casting process as well. The mould was preheated with hot oil flowing through the mould by means of a heating oil pumping unit. An electric resistance furnace was used to melt the ingots of the alloy. The calibrated K-type thermocouples were used to measure the temperature of molten metal and the mould. To transfer molten metal from the furnace to the mould, a ladle made of reinforced fibreglass (RFG) coated with boron nitride was used owing to its non-wetting characteristics with molten metal, at three different pouring temperatures, i.e., 680, 710 and 740 °C. Figure 2 shows the main components of the test equipment used in the present work, along with a schematic illustration of the molten metal pouring stage into the mould cavity.

2.3. Repeatability

The term ‘repeatability’ in fluidity testing refers to the ability of the method to generate similar fluidity results for multiple tests performed under the same conditions. It is the repeatability of results that defines the reliability of any proposed method for practical use at an industrial scale. Typically, the fluidity of metals or alloys is highly sensitive towards temperature, and during experimental testing, this objective cannot be easily achieved, as molten metal continuously losses temperature during solidification. Therefore, ideal fluidity testing of an alloy requires careful monitoring of all process variables, especially the mould and melt temperature, precisely to avoid fluctuations in results repeatability. Therefore, it is necessary to design a proper monitoring path of all the process-related variables along with a standardized operating procedure to avoid any interference in the comparability of results. The operating protocols used in the present work are discussed below:

• All experiments were conducted in an open environment with a temperature of ~20 °C.
• Before inserting the ingots of aluminium alloy into the furnace, adherent dust and greasy materials present on them were thoroughly cleaned by the acetone to avoid possible inclusions in casting.
• The dross was removed from the molten melt surface every time shortly before pouring into the mould.
• The ladle was filled to the same level with molten metal to maintain a constant metallostatic pressure across all the experiments.
• The filling of molten metal was standardized by targeting the same pouring zone in the pouring basin.
• The pouring of molten metal was performed with constant pouring velocity and the same pouring angle to ensure uniformity across all sets of experiments.
• The temperature monitoring of molten metal and mould was performed consistently before conducting each experiment by means of K-type thermocouples.

Following the above-mentioned operative protocols, the implemented operational sequence is given below:

• To perform a series of experiments, aluminium ingots of ~40 kg were melted in an electric resistance furnace with a maximum capacity of 50 kg.
• To reduce the chilling effect, hot oil was continuously circulating through the mould (cope and drag) using a heating oil pumping unit before pouring the molten metal into the mould cavity to have a homogeneous and stable temperature in the mould cavity. The thermocouple to measure the mould temperature was placed exactly at the inlet of the mould where hot oil entered from the connector tube to the mould. The recorded mould temperature was 105 ± 3 °C.
• The interior surfaces of cope and drag were coated with boron nitride to suppress heat transfer from the molten metal to the mould. This coating also helps to control heat transfer to achieve better directional solidification and ensure correct filling of the cavity.
• To avoid any possible levelling error, drag was placed on a perfectly horizontal surface. After that, the cope was placed over the drag by aligning the side holes with those on the drag until the bolts could pass through each other. After aligning, the nuts were screwed tightly to have two halves firmly tight together.
• The ladle was preheated for 5 min and was held in the molten metal bath for 15 s before pouring the molten metal into the mould.
• To guarantee stable and homogeneous filling, molten metal was poured at the identified pouring zone in the pouring basin from the same side to prevent any difference in the filling dynamics of molten metal for all sets of experiments.
• After pouring, a solidification time of ~30 s was applied before opening the mould to remove the casting.

2.4. Numerical Modelling of the Fluidity Test

The behaviour of the molten metal inside the fluidity mould was studied by 3D transient simulations using the commercial software Altair^® Inspire™ Cast (version 2021.2, Altair Engineering, Troy, MI, USA). The boundary conditions for heat transfer were established to simulate the interaction between the molten metal, the mould, and the surrounding environment. Heat transfer between the melt and the mould was considered ideal, ensuring effective thermal exchange based on the predefined temperatures. Additionally, heat dissipation to the surrounding environment is accounted for through convection, using an Environmental Heat Transfer Coefficient (HTC) set at 20 W/m²·K. The objective of the simulation was to analyze the metal flow front during filling, as well as its temperature distribution and duration of both the filling and the solidification processes.

In Figure 3, the geometry simulated, as well as the mesh details, are shown. The operating conditions used for experimental testing were taken as a reference for numerical simulations. With the objective to achieve a high correlation between the numerical simulation and the experimental tests, it was assumed that the molten metal loses approximately 10 °C during the molten metal transfer from the furnace to the pouring basin, and thus the implemented temperatures in the simulations were 670, 700, and 730 °C. However, for comparison, all the cases were referred to as 680, 710, and 740 °C. The mould temperature was set at 105 °C. The distance between the ladle and pouring basin (H) was set to 0.09 m, which gives the pouring velocity ($v$) of 1.33 m/s using the following equation:

$$H={\displaystyle \frac{v^2}{2g}}$$

(1)

where g is the acceleration due to gravity (9.8 m/s²). Additionally, the mesh has been conducted considering 588k tetrahedral elements. With these conditions, simulations were performed where the filling and the solidification process of the part on the fluidity mould were analyzed.

3. Results and Discussion

Following above designed protocols and operational sequence, seven experiments were performed at each pouring temperature. The average value of all the tests carried out at each condition, as well as the fluidity variation obtained in each channel at different pouring temperatures, is given in Figure 4.

The results clearly show that with increasing pouring temperature, the fluidity of a given alloy increases. At higher pouring temperatures, i.e., higher melt superheat slows down the nucleation and growth of fine grains at the tip of flowing metal in the testing channels. This higher melt superheat extends the molten metal lifespan allowing the molten metal to flow for a longer period, and hence higher fluidity is achieved. The increase in fluidity as a function of pouring temperature has also been stated by multiple investigators [39,40]. For a section thickness of 1 mm, no fluidity was measured for any of the investigated pouring temperatures because the alloy was unable to overcome the surface tension in this thickness value. However, all other channels, at a 3 mm thickness and above, were able to fill the mould cavity with various lengths. The trend further reveals that fluidity increases as a function of channel thickness for the same pouring temperature. However, anomalous behaviour for the 3 mm channel was observed at pouring temperatures of 680 °C and 710 °C. At a pouring temperature of 680 °C, the fluidity in the channel with a 3 mm thickness was measured higher than the 5 and 7 mm channel thicknesses, while in the 710 °C case, it was higher than the channels with 5, 7 and 9 mm thicknesses. The higher fluidity in the 3 mm channel as compared with 5 and 9 mm at 680 °C and 710 °C may correspond to the placement of this channel near the pouring basin as the molten metal is at its highest temperature there, reducing its viscosity and allowing it to flow more easily. Moreover, the larger metal volume near the pouring basin retains more heat, further enhancing the fluidity and allowing for more effective mould filling before solidification begins. Meanwhile for 740 °C, the normal increasing trend of fluidity with strip thickness was found. The maximum length was observed in the thickest channel, i.e., 11 mm for all pouring temperatures, because of the lowest heat dissipation rate as compared with thin sections. The overall fluidity at each pouring temperature was calculated as the total volume of solidified alloy in all six channels by the following equation [35]:

$$V_f\left(total\right)={\displaystyle \sum _{i=1}^6}\left(A_i\times L_{f_i}\right)$$

(2)

where $V_f\left(total\right)$ is the total volume (mm³), $A_i$ (mm²), and $L_{f_i}$ (mm) are the cross-sectional area and the measured fluidity length in the channel, respectively. The measured fluidity volume for each pouring temperature, along with the standard deviation, is shown in Figure 5. The results showed that fluidity increased by 25% upon increasing the pouring temperature from 680 °C to 710 °C and increased by 45% from 710 °C to 740 °C. The results further reveal that with increasing pouring temperature, the deviation in the fluidity values decreases, which means that there is a higher repeatability at higher pouring temperatures. The higher repeatability at higher pouring temperatures was achieved because the molten metal remains in a liquid state for a longer period, reducing variability in mould filling and solidification. Moreover, higher temperatures make the process less sensitive to minor variations in pouring conditions and hence lead to more stable and uniform castings. In contrast, lower temperatures increase the risk of premature solidification, inconsistent flow, and greater variation between castings, thus reducing repeatability.

The molten metal behaviour at all pouring temperatures was compared with the numerical simulation results. The temperature predicted by the numerical model at the beginning of the solidification stage at the end of each strip is summarized in Figure 6. The simulations predict the complete filling of all the mould strips at all the pouring temperatures investigated. However, considering the solidification phenomena, it was observed that in all the cases for 1 mm of strip thickness, the temperature at the beginning of the solidification stage fell below the liquidus temperature of the alloy. This value indicates that, at the end of the filling stage, the molten metal should not have been able to flow correctly in the strip. The liquidus temperature measured experimentally for the alloy was 599.8 °C. If we cut the chart at this liquidus temperature, the value obtained by the simulation results at the beginning of the solidification process shows similar behaviour to the strip lengths measured in real experimentation, as shown in Figure 4.

In casting, the fluidity of the metal is intrinsically related to the time when the metal starts to solidify. From the simulations carried out at different pouring temperatures, a time when the solidification begins has been calculated and is shown in

Table 2. Time to the start of solidification in each strip at the studied pouring temperature (numerical simulation).

Channel	Thickness	680 °C	710 °C	740 °C
1	9 mm	0.439 s	0.502 s	0.628 s
2	7 mm	0.402 s	0.471 s	0.543 s
3	3 mm	0.239 s	0.336 s	0.406 s
4	1 mm	0.106 s	0.107 s	0.169 s
5	5 mm	0.357 s	0.396 s	0.483 s
6	11 mm	0.454 s	0.540 s	0.656 s

.

As expected, the higher the pouring temperature, the longer the molten metal takes to start the solidification process. At the same pouring temperature, the strip with a thicker cross-section area has a low heat dissipation rate, and thus it requires more time to solidify as compared with thin sections. However, an interesting observation was obtained when the average solidification time was plotted against the pouring temperature (as shown in Figure 7). It can be seen that, as in the experimental fluidity tests, the value for the intermediate temperature (710 °C) falls below a straight tendency line, which suggests that the simulation, despite differences in the predicted global filling pattern and the temperatures reached during the filling, reproduces very well the trend observed experimentally in the fluidity tests. Therefore, the relationship between fluidity and temperature observed experimentally replicates well the theoretical mechanisms behind the tests, as a similar trend, which seems to be non-linear, is predicted between the initial solidification time and the pouring temperature by the numerical model.

4. Conclusions

In this work, the development of a multi-channel testing methodology to evaluate the fluidity of aluminium foundry alloys for thin wall casting applications was presented and discussed. The following conclusions are drawn on the basis of the achieved results:

• The fluidity increases as a function of channel thickness. In all studied conditions, a channel with 11 mm of thickness shows the highest fluidity as compared with other sections because of a lower heat dissipation rate. At the same time, no fluidity was measured for 1 mm of channel thickness because the alloy was unable to overcome the surface tension in this thickness value.
• The fluidity of the given alloy increases with the pouring temperature. The fluidity increases from 25% to 45% when the pouring temperature is increased from 680 °C to 710 °C and from 710 °C to 740 °C respectively, because of the greater super cooling and nucleation time.
• With an increase in pouring temperature, the deviation in fluidity is reduced, i.e., greater repeatability is achieved.
• Numerical simulation shows a good correlation with real experimentation. The flow length increases with the pouring temperature. Along with that, an incremental trend for increasing the flow length with increasing the section thickness was also observed.
• The designed methodology, if operated with control and accuracy, provides a reliable pathway for foundries to determine the fluidity of aluminium alloys for various industrial applications.

Studying and analysing the impact of mould surface roughness on heat transfer and solidification in thin wall castings through experimental and numerical simulation may be of interest to future studies.