**3.3.1 Microstructural observations**

Fig. 15 shows the microstructure of artificial aged wheels, which were water quenched at 45 and 95°C. The size and shape of the eutectic Si particles were not influenced by the quenching condition used in the present work. The different quenching media influenced probably the Mg2Si and Si precipitates in the α-Al matrix obtained by subsequent artificial ageing. Detailed TEM investigations on A356 alloy, reported elsewhere (Zhang & Zheng, 1996), revealed that, at the peak-aged condition and with a water quench at 25°C, the α-Al matrix consists of a large number of needle-shaped and coherent β″-Mg2Si precipitates. The size of the precipitates is approximately 3 to 4 nm in diameter and 10 to 20 in length. With a

Fig. 15. Microstructure in the hub of the wheels; the micrographs refer to artificial aged A356 alloy solubilised at 540°C for 6 h and water quenched at (a) 45 and (b) 95°C

Optimizing the Heat Treatment Process of Cast Aluminium Alloys 213

temperatures lead to different amount of distortions. Fig. 17 shows the wheel distortion after quenching as a function of the water temperature. By increasing the water temperature, the amount of distortion is reduced; for instance, water at 95°C produces an overall distortion of 1.1 mm, while the wheel distortion is increased up to 1.9 mm with water quenching at 45°C. In the present work, the relationship between the overall casting distortion after quenching and the water temperature has been estimated by linear regression analysis (Fig. 17). The

where T is the water temperature in °C. The regression analysis leads to R2 equal to 0.96.

**40 50 60 70 80 90 100 Water temperature (°C)**

Fig. 17. Wheel distortion after quenching as a function of water quenching temperature;

It has to be mentioned, that the distortion measurements after quenching was carried out on a batch of wheels that were previously cooled in water at 30°C at the exit of the LPDC machine. As previously seen, this operation produces an average distortion of about 1.1 mm. Thus, the "real" distortion caused by water quenching ε′ was calculated by removing the effect of post-cast cooling (Fig. 18). Again, the wheel distortion progressively reduces by increasing the water temperature, and with a temperature higher than 80°C is approximately zero. This behaviour is explained considering the cooling history and the heat transfer condition of an isothermal mass being quenched from a high initial temperature (solution temperature) in a stagnant bath of liquid. Bath quenching starts with a relatively slow rate of cooling, apparently due to a very rapid development of a thin vapour layer which prevents from the contact of "new" water. The film boiling regime persists from elevated surface temperatures down to a lower temperature limit commonly

**0**

standard deviations are given as error bars

**0.5**

**1**

**1.5**

**Distortion,** ε**t (mm)**

**2**

**2.5**

ε<sup>t</sup> =− ⋅ + 0.022 T 3.101 (5)

**Linear fit (R2=0.97)**

distortion εt can be described according to the following regression model:

water quench at 60°C, Zhang and Zheng observed how the density of the precipitates decreases and the size of the precipitates increases slightly; at the same time a significant number of fine Si precipitates resulting from precipitation of excess Si could be observed in the α-Al matrix.

With a slow quenching in air, very different precipitation features are normally evidenced. By air quenching, the material remains at high temperatures for a longer period, which enhances the diffusion of silicon and magnesium. Besides a high density of fine β″-Mg2Si precipitates, the α-Al matrix also contained a large number of areas with coarse rods β′-Mg2Si grouped parallel to each other (Zhang & Zheng, 1996). While the first precipitates have an average size approximately 2 to 3 nm in diameter and around 40 nm in length, the latter show an average size ~15 nm in diameter and 300 nm in length.

Fig. 16. Silicon precipitates within dendrites in A356-T6 wheels that have been slowly quenched in air; arrows indicate the Si particles in the α-Al matrix, as revealed by EDS spectra. Precipitate-free zone (PFZ) is indicated near the eutectic regions

Due to the low Mg content in the present alloy, a high excess Si concentration is present in the α-Al matrix. Assuming the stoichiometric formation of β′-Mg2Si, this alloy concentration should form 0.3 wt.% Mg2Si and an excess of 1 wt.% Si in the alloy, which precipitates as coarse particles within the α-Al matrix (Fig. 16), as revealed by EDS spectra. Further, a clearly visible precipitate-free zone (PFZ) can be seen near the eutectic regions, illustrating that Si has diffused towards existing crystals; such region is marked in Fig. 16.

## **3.3.2 Distortion behaviour of quenched wheels**

The overall distortion εt on 18-inch wheels was measured after quenching in water at different temperature. The different quenching rates obtained using water at different

water quench at 60°C, Zhang and Zheng observed how the density of the precipitates decreases and the size of the precipitates increases slightly; at the same time a significant number of fine Si precipitates resulting from precipitation of excess Si could be observed in

With a slow quenching in air, very different precipitation features are normally evidenced. By air quenching, the material remains at high temperatures for a longer period, which enhances the diffusion of silicon and magnesium. Besides a high density of fine β″-Mg2Si precipitates, the α-Al matrix also contained a large number of areas with coarse rods β′-Mg2Si grouped parallel to each other (Zhang & Zheng, 1996). While the first precipitates have an average size approximately 2 to 3 nm in diameter and around 40 nm in length, the

Fig. 16. Silicon precipitates within dendrites in A356-T6 wheels that have been slowly quenched in air; arrows indicate the Si particles in the α-Al matrix, as revealed by EDS

Due to the low Mg content in the present alloy, a high excess Si concentration is present in the α-Al matrix. Assuming the stoichiometric formation of β′-Mg2Si, this alloy concentration should form 0.3 wt.% Mg2Si and an excess of 1 wt.% Si in the alloy, which precipitates as coarse particles within the α-Al matrix (Fig. 16), as revealed by EDS spectra. Further, a clearly visible precipitate-free zone (PFZ) can be seen near the eutectic regions, illustrating

The overall distortion εt on 18-inch wheels was measured after quenching in water at different temperature. The different quenching rates obtained using water at different

spectra. Precipitate-free zone (PFZ) is indicated near the eutectic regions

that Si has diffused towards existing crystals; such region is marked in Fig. 16.

**3.3.2 Distortion behaviour of quenched wheels** 

latter show an average size ~15 nm in diameter and 300 nm in length.

the α-Al matrix.

temperatures lead to different amount of distortions. Fig. 17 shows the wheel distortion after quenching as a function of the water temperature. By increasing the water temperature, the amount of distortion is reduced; for instance, water at 95°C produces an overall distortion of 1.1 mm, while the wheel distortion is increased up to 1.9 mm with water quenching at 45°C. In the present work, the relationship between the overall casting distortion after quenching and the water temperature has been estimated by linear regression analysis (Fig. 17). The distortion εt can be described according to the following regression model:

$$
\varepsilon\_t = -0.022 \cdot \text{T} + 3.101 \tag{5}
$$

where T is the water temperature in °C. The regression analysis leads to R2 equal to 0.96.

Fig. 17. Wheel distortion after quenching as a function of water quenching temperature; standard deviations are given as error bars

It has to be mentioned, that the distortion measurements after quenching was carried out on a batch of wheels that were previously cooled in water at 30°C at the exit of the LPDC machine. As previously seen, this operation produces an average distortion of about 1.1 mm. Thus, the "real" distortion caused by water quenching ε′ was calculated by removing the effect of post-cast cooling (Fig. 18). Again, the wheel distortion progressively reduces by increasing the water temperature, and with a temperature higher than 80°C is approximately zero. This behaviour is explained considering the cooling history and the heat transfer condition of an isothermal mass being quenched from a high initial temperature (solution temperature) in a stagnant bath of liquid. Bath quenching starts with a relatively slow rate of cooling, apparently due to a very rapid development of a thin vapour layer which prevents from the contact of "new" water. The film boiling regime persists from elevated surface temperatures down to a lower temperature limit commonly

Optimizing the Heat Treatment Process of Cast Aluminium Alloys 215

**0**

temperature of water quenching

basket is immersed in water (Fig. 20).

 Direction of vapour pockets

**0.5**

**1**

**1.5**

**Distortion,** ε **(mm)**

**2**

**1° 2° 3° 4°**

**5°**

**1°**

**3°4°5°**

**1°2° 3°**

**48°C 58°C 67°C 75°C 81°C 86°C 89°C 94°C**

**4°**

**1°**

**4°**

**5° 1°**

**2° 3°**

**4° 5°**

**1° 2° 3° 4° 5°**

**1°, 2°, 3°, 4°, 5° = No. plane in the basket**

Wheel

**1°**

**2°3°**

**4° 5°** **1°2° 3° 4°**

**5°**

Accumulation of vapour pockets

**2°3°**

Fig. 19. Average distortion of the wheels in the five planes of the basket as a function of the

Generally, the first and the last planes of the frame present the extreme values of distortion. This can be explained considering the quenching operation. The wheels at the first planes of the basket are the first to enter in the water bath and their immersion produces a strong water evaporation with the formation of large vapour pockets, which go up toward the bath surface. The amount of vapour increases progressively at the top of the water bath, as the

Fig. 20. Draft of the vapour accumulation at the top of the water bath; as the supporting basket is progressively immersed in water, the wheels produce strong water evaporation

with the formation of large vapour pockets, which go up toward the bath surface

**5°**

**2°**

**2.5**

referred to as the minimum heat flux or Leidenfrost temperature. Below this temperature limit there exists the transition boiling regime, in which the droplets begin to effectively wet the surface resulting in higher heat transfer rates and a faster decrease in the surface temperature. As the surface temperature decreases in the transition boiling regime from the Leidenfrost temperature, the heat transfer rate increases. At the lower temperature boundary of the transition boiling regime, the heat transfer rate reaches a maximum and the temperature of the mass drops rapidly (Liščič et al., 2010; Bernardin et al., 1997).

Fig. 18. Effective wheel distortion ε′ caused by quenching as a function of temperature of water quenching; standard deviations are given as error bars

By using warm water, the Leidenfrost temperature shifts to lower values and the film boiling regime is stable in a greater temperature range. In the range of stable film boiling the temperature falls slowly, almost independent of the bath temperature. Therefore, a uniform cooling is obtained throughout the wheel and the amount of distortion is reduced. Contrary, if the temperature falls soon below the Leidenfrost temperature, the film boiling collapses and the temperature drops rapidly. The higher the Leidenfrost temperature is, that is the sooner the film collapses, the shorter is the total quenching time. Therefore, the 18-inch wheels quenched in water bath at a temperature higher than 80°C keep the initial distortion caused by rapid cooling after casting process.

Even if the non-homogeneous cooling of the casting during quenching remains the main cause of the distortions, another important feature to be considered is the non-homogeneous heat exchange of the batch of wheels inside the water tank. In automotive wheel production, generally, several wheels are contemporary quenched by using a steel basket. In this work, batches of 30 wheels, automatically loaded in a five plane steel frame, are quenched. The different heat transfer conditions created in the water bath influence the distortion behaviour of the wheels in the basket. Fig. 19 shows the average distortion of the wheels at the different planes of the basket as a function of the water quenching temperature.

referred to as the minimum heat flux or Leidenfrost temperature. Below this temperature limit there exists the transition boiling regime, in which the droplets begin to effectively wet the surface resulting in higher heat transfer rates and a faster decrease in the surface temperature. As the surface temperature decreases in the transition boiling regime from the Leidenfrost temperature, the heat transfer rate increases. At the lower temperature boundary of the transition boiling regime, the heat transfer rate reaches a maximum and the

> **40 50 60 70 80 90 100 Water temperature (°C)**

Fig. 18. Effective wheel distortion ε′ caused by quenching as a function of temperature of

By using warm water, the Leidenfrost temperature shifts to lower values and the film boiling regime is stable in a greater temperature range. In the range of stable film boiling the temperature falls slowly, almost independent of the bath temperature. Therefore, a uniform cooling is obtained throughout the wheel and the amount of distortion is reduced. Contrary, if the temperature falls soon below the Leidenfrost temperature, the film boiling collapses and the temperature drops rapidly. The higher the Leidenfrost temperature is, that is the sooner the film collapses, the shorter is the total quenching time. Therefore, the 18-inch wheels quenched in water bath at a temperature higher than 80°C keep the initial distortion

Even if the non-homogeneous cooling of the casting during quenching remains the main cause of the distortions, another important feature to be considered is the non-homogeneous heat exchange of the batch of wheels inside the water tank. In automotive wheel production, generally, several wheels are contemporary quenched by using a steel basket. In this work, batches of 30 wheels, automatically loaded in a five plane steel frame, are quenched. The different heat transfer conditions created in the water bath influence the distortion behaviour of the wheels in the basket. Fig. 19 shows the average distortion of the wheels at

the different planes of the basket as a function of the water quenching temperature.

**Fit Line**

**-1**

caused by rapid cooling after casting process.

water quenching; standard deviations are given as error bars

**-0.5**

**0**

**0.5**

**Distortion,** ε**' (mm)**

**1**

**1.5**

**2**

temperature of the mass drops rapidly (Liščič et al., 2010; Bernardin et al., 1997).

Fig. 19. Average distortion of the wheels in the five planes of the basket as a function of the temperature of water quenching

Generally, the first and the last planes of the frame present the extreme values of distortion. This can be explained considering the quenching operation. The wheels at the first planes of the basket are the first to enter in the water bath and their immersion produces a strong water evaporation with the formation of large vapour pockets, which go up toward the bath surface. The amount of vapour increases progressively at the top of the water bath, as the basket is immersed in water (Fig. 20).

Fig. 20. Draft of the vapour accumulation at the top of the water bath; as the supporting basket is progressively immersed in water, the wheels produce strong water evaporation with the formation of large vapour pockets, which go up toward the bath surface

Optimizing the Heat Treatment Process of Cast Aluminium Alloys 217

may be too great to reverse. Generally, several coats are applied to aluminium wheels to guarantee a suitable corrosion resistance. After each coat the wheels are left inside an air electric furnace for drying at 170 ± 5°C for 1 hour. From the heating curve in Fig. 3, it is observed that it takes approximately 20 minutes to heat the wheels from room temperature to 145°C. Due to slow heating, the coating treatment effect experienced by the wheels during the heating stage is not negligible. Then, the wheels are maintained for 35 minutes in a range of temperature between 145 and 170°C. The temperature and time used in the present work for powder coating activate the diffusion mechanism of the solute atoms, such as Mg and Si, leading to the precipitation of dissolved elements and the coarsening of existing precipitates, i.e. the bake hardening effect. The influence of powder coating cycles on the hardness of T6 heat treated wheels is shown in Fig. 22. The hardness increases progressively after each coating cycles of about 3%. The average hardness of wheels after machining is around 92

**No. coating cycle**

Fig. 22. Effect of coating cycles on hardness of wheels, which were solution treated at 540°C for 6 h, water quenched and aged at 145°C for 4 h; data refer to water quenching at different

In the present work, some process variables, which play a key role in production cycle of wheels have been investigated and improved. An integrated methodology for developing and optimizing the production and the final quality of A356-T6 18-inch wheels, in terms of casting distortion and hardness, has been proposed. This study has focused on examining both the effect of cooling rate on wheel distortion and hardness during the post-cast and quenching steps, and the influence of the solutionizing temperature and time, and the

**0 cycle 1 cycle 2 cycle 3 cycle**

HB, while after 3 coating cycles the hardness increases up to 98 HB.

**80**

temperature

**4. Conclusions** 

**85**

**90**

**75°C 80°C**

**85°C 90°C**

**95°C**

**75°C**

**80°C 85°C**

> **90°C 95°C**

> > **75°C**

**80°C 85°C**

**90°C**

**75°C**

**80°C 85°C 90°C**

**95°C**

**95°C**

**95**

**Hardness (HB5/250/30)**

**100**

**105**

The vapour pockets may collapse on the casting surface and locally change the heat transfer coefficient between the piece and the quenchant by preventing from the contact of "new" water. Once again, a non-homogeneous quenching rate is established throughout the wheel. The wheels at last planes of the basket undergo different quenching conditions than those at the first planes.

The influence of water temperature on hardness of wheels after ageing at 145°C for 4 h is shown in Fig. 21. The different water temperature, in the range between 40 and 95°C, doesn't influence (to some extent) the hardness properties of the A356 alloy, that is the hardness fluctuates slightly around 92 HB. Generally, the hardness of A356 alloy decreases by lowering quench rates. It has been studied that with a quench rate higher than 110°C/s, obtained with water at temperature lower than 60°C, the peak hardness of A356 alloy is not influenced by the quench rate (Zhang & Zheng, 1996); nevertheless, a little difference (~4 HB) occurs by water quenching in the temperature range between 60 and 100°C (Fracasso, 2010). Furthermore, the time to peak hardness increases for extremely slow quench rates (0.5°C/s), while for faster quench rates, above 20°C/s, no shift is seen in the time to the peak. Therefore, by increasing the temperature of water quenching up to 95°C, the target hardness of the wheels after a complete T6 heat treatment is achieved and the wheel distortion is reduced.

Fig. 21. Brinell hardness measured throughout the wheel as a function of the different temperature of water quenching; standard deviations are given as error bars. Data refer to wheels solution treated at 540°C for 6 h and aged at 145°C for 4 h

#### **3.4 Powder coating**

Most aluminium wheels are clear coated for corrosion resistance and aesthetic appearance. Unprotected aluminium wheels quickly corrode and pit when exposed to road salt and excessive moisture. If the corrosion continues unchecked for too long, the cosmetic damage

The vapour pockets may collapse on the casting surface and locally change the heat transfer coefficient between the piece and the quenchant by preventing from the contact of "new" water. Once again, a non-homogeneous quenching rate is established throughout the wheel. The wheels at last planes of the basket undergo different quenching conditions than those at

The influence of water temperature on hardness of wheels after ageing at 145°C for 4 h is shown in Fig. 21. The different water temperature, in the range between 40 and 95°C, doesn't influence (to some extent) the hardness properties of the A356 alloy, that is the hardness fluctuates slightly around 92 HB. Generally, the hardness of A356 alloy decreases by lowering quench rates. It has been studied that with a quench rate higher than 110°C/s, obtained with water at temperature lower than 60°C, the peak hardness of A356 alloy is not influenced by the quench rate (Zhang & Zheng, 1996); nevertheless, a little difference (~4 HB) occurs by water quenching in the temperature range between 60 and 100°C (Fracasso, 2010). Furthermore, the time to peak hardness increases for extremely slow quench rates (0.5°C/s), while for faster quench rates, above 20°C/s, no shift is seen in the time to the peak. Therefore, by increasing the temperature of water quenching up to 95°C, the target hardness of the wheels after a complete T6 heat treatment is achieved and the

> **40 50 60 70 80 90 100 Water temperature (°C)**

Most aluminium wheels are clear coated for corrosion resistance and aesthetic appearance. Unprotected aluminium wheels quickly corrode and pit when exposed to road salt and excessive moisture. If the corrosion continues unchecked for too long, the cosmetic damage

Fig. 21. Brinell hardness measured throughout the wheel as a function of the different temperature of water quenching; standard deviations are given as error bars. Data refer to

wheels solution treated at 540°C for 6 h and aged at 145°C for 4 h

the first planes.

wheel distortion is reduced.

**70**

**3.4 Powder coating** 

**75**

**80**

**85**

**Hardness (HB5/250/30)**

**90**

**95**

**100**

may be too great to reverse. Generally, several coats are applied to aluminium wheels to guarantee a suitable corrosion resistance. After each coat the wheels are left inside an air electric furnace for drying at 170 ± 5°C for 1 hour. From the heating curve in Fig. 3, it is observed that it takes approximately 20 minutes to heat the wheels from room temperature to 145°C. Due to slow heating, the coating treatment effect experienced by the wheels during the heating stage is not negligible. Then, the wheels are maintained for 35 minutes in a range of temperature between 145 and 170°C. The temperature and time used in the present work for powder coating activate the diffusion mechanism of the solute atoms, such as Mg and Si, leading to the precipitation of dissolved elements and the coarsening of existing precipitates, i.e. the bake hardening effect. The influence of powder coating cycles on the hardness of T6 heat treated wheels is shown in Fig. 22. The hardness increases progressively after each coating cycles of about 3%. The average hardness of wheels after machining is around 92 HB, while after 3 coating cycles the hardness increases up to 98 HB.

Fig. 22. Effect of coating cycles on hardness of wheels, which were solution treated at 540°C for 6 h, water quenched and aged at 145°C for 4 h; data refer to water quenching at different temperature
