Abstract: CAE simulation software was used to analyze sequence and temperature field of aluminum alloy transmission housings during filling and solidification processes. Locations prone to defects such as air entrapment and shrinkage porosity were observed and their causes analyzed. Based on simulation results, relevant improvements to die-casting process were proposed, namely, increasing number of runners in gating system and optimizing overflow system design. Finally, actual production and X-ray inspection showed that defects were improved, verifying feasibility of proposed solution and providing a reference for casting qualified parts.
With rapid development of automotive lightweighting, lightweight alloy materials are increasingly being used in component materials. Aluminum alloy, as a lightweight alloy, meets needs of automotive lightweighting, possessing characteristics such as low density and good friction performance, and is widely used in production of castings such as automotive transmission housings. Automotive transmission housings are generally produced by die-casting aluminum alloy blanks followed by subsequent machining. In actual production, defects such as porosity, cold shuts, and cracks often affect their performance.
Simulation analysis is the most effective auxiliary tool for casting process and mold design. It can accurately and efficiently simulate real casting process, thereby shortening mold and process development cycle, reducing design and production costs. This study focuses on shell castings, which have complex structures and uneven wall thicknesses. During filling and solidification processes, defects such as porosity, shrinkage cavities, and shrinkage holes may occur, making these areas prone to cracking. Chen Ziye et al. used casting simulation software to simulate and analyze temperature field changes during filling and solidification processes of castings, analyzed location and causes of defects, and optimized process to improve casting quality. Luo Wei et al. used simulation software to simulate automotive lamp brackets, analyzed defects that occurred during casting process, and proposed corresponding solutions.
This study primarily utilizes simulation software Anycasting to conduct computer simulations of filling and solidification processes of an aluminum alloy gearbox housing. Aim is to optimize die-casting process, observe flow field, temperature field, defect distribution of molten aluminum alloy during filling and solidification, predict location of defects, analyze their causes, and thus optimize relevant processes. Finally, through trial production of final casting, quality level of product is improved and production cycle is shortened.

Figure 1 shows three-dimensional model of gearbox housing die-casting. Die-casting material is ADC12 aluminum alloy, and mold material is H13 steel. Model weighs approximately 4.02 kg, with outline dimensions of 409.9 mm * 283.2 mm * 81.9 mm and a volume of approximately 1,487,211.88 mm³. As shown in Figure 2, wall thickness of this die-cast part is uneven, with a maximum wall thickness of 19.78 mm and an average wall thickness of 5.68 mm. From front and back views, it can be seen that it is a complex type of component with numerous internal reinforcing ribs and holes. Gearbox housing uses a top-pour gating system, with ingate located at the bottom of casting. Overflow grooves are also provided in key areas for venting and storing cold material.

 

Fig. 1 Three-dimensional model of casting
Casting material is ADC12 aluminum alloy, with a liquidus temperature of 580 ℃ and a solidus temperature of 515 ℃. ADC12 aluminum alloy has excellent casting performance, high yield strength, and good creep resistance at high temperatures, making it widely used in casting industry. Chemical composition of ADC12 alloy is shown in Table 1, and mold material is H13 steel, with its thermophysical parameters shown in Table 2.

 

Fig. 2 Wall thickness diagram of casting

3D model of gearbox housing was converted to STL format and imported into Any PRE software. Software's uniform mesh generation function was used to perform mesh generation. To improve meshing time and minimize computation while ensuring no part of solid is missed, a global meshing approach is adopted, resulting in a total of 4,148,144 meshes. Meshing model is shown in Figure 3. As can be seen from figure, meshing is reasonable, with no discontinuities.

Table 1 Chemical composition of the ADC12 alloy wB /%

Table 2 Thermophysical properties of ADC12 alloy and H13 steel

In casting process, selection of parting surface has a significant impact on quality, manufacturing cost, efficiency of casting. Parting surface of die casting mold is interface between fixed mold and moving mold. Parting surface should ideally be selected on a large contour surface of casting so that die casting remains on moving mold after mold opening, simplifying mold structure, ensuring dimensional and surface quality of die casting. Secondly, parting surface should be oriented in casting direction, which helps ensure smooth flow of molten metal during casting, reduces generation of porosity and inclusions. The largest projected surface of this shell is its bottom. Considering casting structure and mold feasibility, bottom surface of shell is chosen as parting surface, as shown in Figure 4. This parting facilitates core placement, casting removal, cleaning, makes gating and overflow channels easy to remove, ensuring machining quality of casting.

 

Fig. 3 Grid division of die casting

Gating system generally consists of three parts: sprue, runner, and ingate. Rationality of gating system design has a significant impact on filling process. For complex castings such as gearbox housings, ingate design involves many issues, especially ensuring ingate thickness is appropriate for rapid filling. Formula for calculating cross-sectional area of ingate is as follows:

 

Where: Ag is cross-sectional area of ingate, mm2; V is volume of casting, mm3; v is filling speed, mm/s; t is filling time, s. Volume of gearbox housing measured with an nx12.0 is 1487211.88 mm3. V is calculated based on a filling speed of 40000 mm/s, and t is calculated based on 0.1 s. Calculated cross-sectional area of ingate, Ag, is 372 mm2. Since average thickness of gearbox housing is 5.68 mm, and a smaller ingate thickness is chosen to facilitate removal of gating system later, ingate thickness is selected as 3 mm. Sprue is connecting channel from end of sprue to the front of ingate. Shape, position, direction, and size of ingate determine its structural shape. Thickness of sprue is calculated using formula (2): D = (5~8)h (2)
Where: D is sprue thickness, mm; h is ingate thickness, mm; D = 16 mm is taken. To make casting easier to remove from mold, draft angle of sprue is generally taken as 10°~15°, here 15° is chosen.

 

Fig. 4 Determination of parting surface
This study uses a Lijing DCC1600T horizontal cold chamber die-casting machine, with a chamber diameter of 110 mm and a sprue thickness of 30 mm.

Main process parameters for die casting include pouring temperature, filling speed, and mold preheating temperature, all of which affect final quality of casting. Based on structural characteristics of die casting, pouring temperature is selected as 670 ℃. Mold preheating ensures good fluidity of molten metal during filling, thus facilitating better filling. Therefore, a reasonable mold preheating temperature must be selected during production; selected temperature is 200 ℃. A reasonable filling speed can improve quality of die casting, but filling speeds are divided into high and low speeds. Therefore, low speed is set to 0.25 m/s, and high speed to 3 m/s. Specific parameters are shown in Table 3.

Table 3 Selection of process parameters

Figure 5 shows simulation results of filling time for die casting. It can be seen that the total filling time is 0.4087 s. Temperature field of molten metal during filling process is shown in Figure 6. During filling process, temperature of molten metal is always higher than liquidus temperature of 580 ℃, thus ensuring good fluidity of molten metal inside mold cavity. As shown in Figure 6a, at t=0.258 2 s, molten metal, under action of punch, enters sprue and runner from gate, then flows to ingate from two main branches. As shown in Figure 6b, at t=0.354 1 s, molten metal in two middle runners has already reached ingate and begun to fill cavity. As shown in Figure 6c, at t=0.365 2 s, during filling process, filling speed in the middle area of cavity is significantly faster than that on both sides, which leads to phenomenon of material flow encirclement, easily forming trapped air and air entrapment. As shown in Figure 6e, at t=0.405 7 s, except for overflow channels on both sides of casting and end cylindrical area and overflow channels, the rest has been filled. However, during filling process, gas in cavity is pushed in filling direction by molten metal. Therefore, at junction and end of filling, material flow reverses, resulting in trapped air or severe cold material and cold shut in this area, easily producing defects such as porosity. At t = 0.4087 s, molten metal had filled the entire cavity of casting and entered overflow groove.

 

Fig. 5 Distribution of filling time

Figure 7 shows solidification time of gearbox housing. The total solidification process of casting took 124.1316 s. Figure shows that middle groove of casting solidified last, potentially leading to shrinkage porosity and other defects. Temperature field of casting at different times during solidification is shown in Figure 8.

 

Fig. 6 Filling process of liquid metal at different times

 

Fig. 7 Solidification time
As shown in Figure 8a, temperature of molten metal did not change significantly during filling process and remained above liquidus. Therefore, local solidification did not occur during filling process. At t = 17.7147 s, overflow groove and thinner areas at the end of casting had already begun to solidify. At t = 21.8076 s, as shown in Figure 8d, temperature at edges and thin-walled areas of casting decreases significantly, while temperature in the middle part of casting remains high. Molten metal cannot be fully filled, and after solidification in low-temperature region, molten metal in this area remains in a solid-liquid coexistence state, forming an isolated liquid phase region, which easily leads to shrinkage porosity and shrinkage cavities. As shown in Figure 8e, at t = 52.0953 s, except for central grooved area which has not yet solidified, the rest of casting has solidified. This is because wall thickness is relatively thick, and there is a risk of continued cooling after product ejects from mold, resulting in surface shrinkage or formation of isolated liquid phase regions internally, potentially leading to shrinkage cavities.

 

Fig. 8 Distribution of temperature field of casting in solidification

Simulation showed that initial filling process of molten alloy was relatively smooth. However, during cavity filling process, some air entrapment and cold shuts occurred, leading to porosity. During solidification, isolated liquid phase zones existed in thick-walled and grooved areas of casting, resulting in shrinkage porosity and shrinkage cavities. To address problems caused by thick walls, a runner was added to each side of casting to accelerate filling speed in these areas and reduce air entrapment. Simultaneously, several overflow channels and an exhaust system were added to casting to prevent localized air trapping, to better remove gas from cavity. Furthermore, by adjusting position and number of gates, flow direction and speed of molten metal were altered, thereby shortening solidification time of casting grooves, reducing shrinkage porosity and shrinkage cavities. Figure 9 shows a schematic diagram of optimized scheme. Two newly added runners on both sides were used to better collect cold material and gas, overflow channels and an exhaust system were added to grooved area at tail of casting.

 

Fig. 9 Schematic diagram of optimization scheme

 

Fig. 10 Optimized filling process

 

Fig. 11 Optimized solidification time

 

Fig. 12 Optimized solidification temperature

 

Fig. 13 Position of section X of casting
Based on above optimization scheme, filling and solidification analysis of optimized product was performed. As shown in Fig. 10, molten metal near ingate is well filled with no obvious air entrapment. In original scheme, air entrapment occurred at the end of casting's circular hole. However, optimized casting has an overflow groove and venting system added at the end to collect cold material and gas, thus solving air entrapment problem at the end. As shown in Fig. 11 and Fig. 12, solidification time of optimized casting is 121.8720 s, which is less than time before optimization. Main reason is that in pretreatment process, spraying time was extended by 2 s, which has a better effect on reducing mold temperature, thereby improving solidification efficiency of molten metal in mold and shortening solidification time of casting. Compared to casting before optimization, area with higher temperature in central groove is significantly reduced, thus lowering probability of shrinkage porosity and shrinkage cavities in this area. As shown in Figure 13, casting is sectioned along direction of arrow X. Defects at section X before and after optimization are shown in Figure 14. Comparing two, it is clear that defects in groove are significantly reduced after optimization, except for some cylindrical holes and overflow channels. However, these defects near gating system are negligible because they will be removed in later processing. Overall, optimized solution reduces risk of air entrapment, shrinkage porosity, shrinkage cavities, and improves production efficiency.

 

Fig. 14 Comparison of defect of X-section before and after optimization

To further verify feasibility of process optimization solution, optimized casting solution was tested on a 16,000 kN Lijing die-casting machine. Die-cast product is shown in Figure 15. Verification showed that rear housing of gearbox had a good surface appearance, without defects such as porosity or cold shuts. X-ray inspection of casting was then performed, and results are shown in Figure 16. Internal structure was good, and no obvious shrinkage porosity or shrinkage cavities were found.

 

Fig. 15 Physical drawing of casting

(1) To effectively address defect problem, number of runners in gating system was increased and overflow system was optimized. Through mold flow analysis, quality of optimized castings was improved, defect rate was effectively reduced, and mass production of products could be achieved, providing a reference for actual production.
(2) To address problem of easy defects and low production efficiency in die-casting process of gearbox housings, numerical analysis of castings was conducted using AnyCasting simulation. Comparison of simulation results with X-ray inspection of castings showed a high degree of consistency, indicating that software can effectively predict location of potential defects in castings during production.

 

Fig. 16 X-ray inspection of casting