Brake drums are critical safety components in the braking systems of heavy-duty trucks, and their quality directly impacts vehicle operational safety and human life. Gray cast iron has long been the preferred material for brake drums due to its excellent thermal conductivity, damping capacity, wear resistance, good castability, and low cost. However, with the increasing load capacity of heavy-duty trucks, especially during prolonged braking on long downhill sections in mountainous areas, the temperature of brake drums often reaches 200–300 ℃. To restore braking capability, water is frequently sprayed onto the drums, and under such repeated heating–quenching conditions, traditional gray cast iron brake drums are highly susceptible to cracking and failure. Therefore, developing or modifying gray cast iron compositions to cast high-performance brake drums is of great significance for extending service life and reducing transportation safety risks.
Alloying methods to adjust the composition of gray cast iron have been shown to improve the mechanical properties of brake drum materials, thereby enhancing their failure resistance and prolonging their service life. However, changes in composition due to alloying lead to alterations in material properties, which in turn affect the filling and solidification behavior during the casting process, potentially causing defects in the workpiece. Traditionally, reducing casting defects has relied on empirical trial-and-error methods, which are costly, time-consuming, and cannot guarantee quality. In recent years, computer-aided engineering (CAE) technology has become increasingly mature in the foundry field. Through numerical simulation, it is possible to analyze the evolution of various field quantities during the filling and solidification processes, thereby efficiently guiding production practices.
In this study, we focus on a modified gray cast iron (MGCI) brake drum developed by a casting enterprise through alloying adjustments to HT250 to improve service performance. We investigate the filling and solidification behavior of the modified material. Using JMatPro software, we calculate the property parameters of the modified material and correct the material model. Based on Procast software, we establish a finite element model for the casting of the MGCI brake drum to predict sand casting defects. The reliability of the model is verified through process experiments. Furthermore, based on this finite element model, we study the effects of pouring temperature and pouring time on the shrinkage porosity of the brake drum, aiming to determine the optimal casting parameters and eliminate sand casting defects.
| Material | C | Si | Mn | S | P | Cr | Ni | Cu | Ti | B | Fe |
|---|---|---|---|---|---|---|---|---|---|---|---|
| HT250 | 3.421 | 1.857 | 0.720 | 0.080 | 0.050 | 0.302 | 0.064 | 0.277 | 0.028 | 0.001 | Bal. |
| MGCI | 3.378 | 1.907 | 0.770 | 0.078 | 0.053 | 0.317 | 0.024 | 0.223 | 0.122 | 0.026 | Bal. |
Material Property Calculations and Finite Element Modeling
In the modified gray cast iron, the Ti and B contents were increased to 0.10 wt.% and 0.03 wt.%, respectively, while the Ni and Cu contents were reduced, aiming to improve the wear resistance and thermal fatigue properties. The raw materials included pig iron, steel scrap, foundry returns, carburizers, and alloy additives, which were melted in a medium-frequency induction furnace. The solidification of gray cast iron involves the precipitation of multiple phases such as graphite and austenite, along with finite diffusion of solute elements in the solid phase. When calculating the thermophysical and fluid dynamic properties of the modified gray cast iron in the temperature range of 20–1500 ℃ using JMatPro software, we selected the Back Diffusion model, which accounts for both equilibrium and non-equilibrium solidification, as the solid diffusion model. Comparing the calculated results with those of the original HT250, significant changes were observed in the solid fraction and density of the modified material.
The solidus and liquidus temperatures of the modified gray cast iron decreased from 1078 ℃ and 1198 ℃ to 1072 ℃ and 1190 ℃, respectively. Within the solid–liquid coexistence temperature range, the variation of solid fraction with temperature became more gradual, especially when the solid fraction approached approximately 0.7, where the solidification rate was notably slower. This is because the increased Ti content introduces more TiC particles that act as nucleation substrates for austenite, refining the grains and promoting nucleation, which shifts the solidification process to lower temperatures. Meanwhile, the addition of B significantly reduces the eutectic reaction temperature, further lowering both the solidus and liquidus. The reduction in Ni and Cu, which have higher melting points, also contributes to the decrease in solidus temperature. The density of the modified gray cast iron was slightly lower, and the difference became more pronounced at low temperatures, indicating a smaller volumetric shrinkage during solidification from the melt to room temperature.
| Property | HT250 | MGCI | Change |
|---|---|---|---|
| Solidus Temperature (℃) | 1078 | 1072 | −6 |
| Liquidus Temperature (℃) | 1198 | 1190 | −8 |
| Density at 20 ℃ (g/cm³) | 7.15 | 7.08 | −0.07 |
| Density at 1200 ℃ (g/cm³) | 6.82 | 6.78 | −0.04 |
| Thermal Conductivity at 500 ℃ (W/m·K) | 42.5 | 41.8 | −0.7 |
The geometric model of the brake drum component has an inner diameter of 320 mm, an outer diameter of 335 mm, and a height of 190 mm, with a mass of 25 kg. The wall thicknesses of the drum body and the bowl mouth are approximately 15 mm and 21 mm, respectively. The casting process采用的是 sand mold casting, designed as a one-mold-two-parts arrangement. The sand mold size is 1200 mm × 700 mm × 450 mm, with the brake drum bowl mouth facing downward, and the molten metal enters from the lower side gate. Since the geometric features of the overall model are mainly concentrated in the casting, especially at the gate, and the other parts are relatively regular and uniform, we采用 a tetrahedral progressive meshing approach to discretize the model. Smaller mesh sizes of 1–3 mm were used for the brake drum, gating system, and riser, while coarser mesh sizes of 10 mm were used for the outermost layer of the sand mold. The transition factor between each layer of mesh from the inner wall to the outer wall of the sand mold was set to 1.1. To ensure mesh quality, all elements were required to have a shape factor greater than or equal to 2. The finite element model comprised 198,506 surface elements, 4,872,325 volume elements, and 833,831 nodes.
Based on actual process conditions, the initial pouring temperature was set to 1350 ℃, and the pouring time was 30 s. During the filling and solidification processes, heat exchange occurs between the melt and the sand mold, as well as between the sand mold and the air. The heat transfer between different parts of the same material was设置为 EQUIV, and the heat transfer type between the casting and the sand mold was设置为 COINC, with a heat transfer coefficient of 500 W/(m²·K). The main heat transfer主体 was the casting, and the convective heat transfer coefficient between the outer surface of the sand mold and the air at 20 ℃ was设置为 20 W/(m²·K). To ensure the accuracy of the numerical simulation, the Flow1 fluid solver mode was selected, and the MicroStructure and New-APM shrinkage porosity sub-modules were enabled.
Analysis of Mold Filling and Solidification Processes
During the filling process, the molten metal enters from the ingate and fills the mold from bottom to top in a sequential manner. The melt rises smoothly without any分流 or convergence of streams, indicating that the design of the gating system is reasonable and the tendency for gas entrapment and inclusions is low. At 28 s, the filling is essentially complete, with the temperature of the metal at the ingate and riser being 1341 ℃ and 1272 ℃, respectively, which are higher than the liquidus temperature of 1190 ℃ of the modified material, suggesting that the melt still has some feeding capacity. The filling time distribution across different parts of the brake drum shows that the sprue and runner are幾乎 completely filled within about 0.8 s. At the beginning of pouring, the molten metal quickly fills the gating system. In the horizontal direction, the filling completion time of each part of the brake drum is幾乎 consistent, and the gradient in the vertical direction is uniform. The bowl mouth ribs near the gating system are filled first, while the upper flange of the brake drum farthest from the gating system is filled last. This indicates that the melt level rises at a constant speed and flows smoothly during the filling process, suggesting that the design of the gating system and the matching of pouring time are reasonable, avoiding the formation of gas inclusions and turbulent flow.
During solidification, the bottom ribs and top flange of the brake drum begin to solidify first, followed by the middle drum wall, while the transition area between the drum wall and the ribs, where the wall thickness changes abruptly, solidifies more slowly and is a potential hot spot region. Overall, although the solidification generally follows the principle of sequential solidification from the workpiece to the riser and from the workpiece to the gate, the runner begins to solidify before the middle and lower parts of the brake drum workpiece. This leads to the formation of an isolated liquid phase region at the rib near the ingate at the final stage of solidification, which cannot receive feeding from the already solidified gate. This is because the upper end of the brake drum is close to the outer wall of the sand mold, and although affected by the heat source of the riser, it still dissipates heat relatively quickly. The lower ribs of the brake drum have the largest wall thickness and are located at the parting surface, in the middle of the entire sand mold, far from the outer wall of the sand mold, and close to the heat source of the gating system, resulting in slower heat dissipation.
The solidification time distribution shows that the riser at the upper end of the brake drum and the area near the ingate at the lower end completely solidify only after about 670 s, which is significantly longer than other positions, failing to meet the expected principle of sequential solidification from the workpiece to the riser/gate. The bottom ribs and top flange of the brake drum solidify first, followed by the middle drum wall, with the overall solidification time of the brake drum ranging from 480 s to 590 s. The gating system is the earliest part of the entire casting to solidify. The sprue completely solidifies first at around 280 s, and the solidification time of each part of the runner increases from 220 s to 550 s as the distance from the sprue decreases, indicating that the gating system fails to achieve the desired feeding effect.
The shrinkage porosity distribution indicates that severe shrinkage defects appear on the ribs of the brake drum near the ingate, which corresponds to the solid fraction distribution at the final stage of solidification and the solidification time. This suggests that an isolated liquid phase region exists near the interface between the workpiece and the ingate at the late stage of solidification. During solidification shrinkage, this region cannot be fed from the surrounding已经完全 solidified parts, leading to relatively severe shrinkage defects. Additionally, there are relatively minor pores in the drum wall. The average porosity of the defective mesh elements is 3.53%. From the shrinkage results, it can be seen that the main problem with the MGCI brake drum casting process is the insufficient feeding capacity of the gating system, which cannot effectively feed the workpiece at the final stage of solidification.
The formation of sand casting defect is closely related to the solidification sequence and the feeding efficiency of the gating system. In sand casting, the defect often manifests as shrinkage porosity in regions that solidify last and are isolated from the feeding path. The key factor leading to this sand casting defect is the premature solidification of the runner and gate relative to the thick sections of the casting, which disrupts the intended progressive solidification pattern. Understanding the mechanism of this sand casting defect is crucial for optimizing the gating system design and process parameters to achieve sound castings. The shrinkage porosity observed in the simulation is a classic example of a sand casting defect that arises when the feeding channel solidifies before the casting region that requires补缩.
Experimental Validation of Sand Casting Defect Prediction
To verify the accuracy of the finite element model for the MGCI brake drum casting, we conducted casting experiments on a static pressure production line under the same process conditions as the numerical simulation. The tapping temperature was 1450 ℃, and chromium iron inoculant was used for ladle inoculation. The comparison between the simulation results and the cast samples shows that the外形 of the MGCI brake drum workpiece is smooth, without quality issues such as gas pores, flash, or sand adhesion, indicating that the filling process meets expectations and corresponds to the stable, gradual filling phenomenon observed in the simulation. Furthermore, significant shrinkage defects were observed on the ribs of the specimen near the ingate, which is basically consistent with the defect location predicted by the simulation. Scanning electron microscopy (SEM) observation of the surface morphology at this location revealed that the pore defects exhibit distinct dendritic morphology, with spherical protrusions distributed on the inner walls. This is because, at the final stage of solidification, the casting undergoes insufficient shrinkage compensation within the isolated liquid phase region, resulting in shrinkage porosity. This sand casting defect is典型的 of inadequate feeding in sand castings with complex geometries and varying wall thicknesses.
The experimental results demonstrate that the established finite element model is capable of accurately simulating the filling and solidification processes of the casting and predicting the shrinkage defects in the brake drum workpiece. The consistency between the simulated and experimental sand casting defect locations validates the reliability of the model for process optimization. The formation of this sand casting defect is primarily attributed to the premature solidification of the gating system relative to the thick rib sections, which disrupts the feeding path and creates isolated liquid pools that contract without replenishment. This type of sand casting defect is commonly encountered in gray iron castings where the solidification range is extended by alloying additions, making the control of solidification sequence and feeding design even more critical.
Process Optimization for Sand Casting Defect Elimination
From the casting simulation and experimental results, it is clear that the MGCI brake drum exhibits shrinkage defects on the ribs near the ingate. To improve the casting quality and eliminate this sand casting defect, we optimized the gating system structure and casting process parameters based on the validated finite element model.
Gating System Optimization
The shrinkage defects appear mainly at the position adjacent to the ingate. During solidification, the final isolated liquid phase region occurs at the hot spot of the workpiece ribs near the ingate because the gating system solidifies earlier than the workpiece and fails to effectively compensate for melt shrinkage. Therefore, based on the existing casting process, we adjusted the gating system structure by increasing the width of the ingate near the sprue and increasing the thickness of the gate well, without affecting the filling and solidification behavior of the workpiece. After the改进, the casting simulation shows that the number of defects is reduced. The overall shrinkage porosity of the mesh elements with pores is about 2.89%, and the location of severe shrinkage defects has shifted from the ribs of the workpiece to the ingate and the bottom platform area of the sprue. The improved gating system can effectively achieve the desired feeding effect, avoiding significant severe shrinkage parts on the workpiece and thus improving the casting quality of the MGCI brake drum. By redesigning the gating system to ensure that the feeding channels remain liquid until the casting sections have fully solidified, we can mitigate the sand casting defect that plagued the original design.
Process Parameter Optimization
Pouring temperature and pouring time significantly influence casting quality. Therefore, considering the actual production process conditions, we selected pouring temperatures of 1350, 1375, 1400, and 1425 ℃, and pouring times of 25, 30, and 35 s, while keeping other parameters constant. A full factorial experiment was conducted based on the finite element method to minimize the shrinkage porosity of the brake drum workpiece.
| Factor | Level 1 | Level 2 | Level 3 | Level 4 |
|---|---|---|---|---|
| Pouring Temperature (℃) | 1350 | 1375 | 1400 | 1425 |
| Pouring Time (s) | 25 | 30 | 35 | — |
When the pouring temperature is in the range of 1350–1375 ℃, the volume of mesh elements with shrinkage defects exceeds 25 cm³ at different pouring times, and the average porosity increases with prolonged pouring time. At a pouring time of 35 s, the porosity reaches 4.17%, while at 25 s, it is 2.19%. This indicates that when casting at this temperature range, the probability of shrinkage defects is high, and the forming quality is poor. This is because, after filling, the melt temperature is较低, the viscosity is high, and the flowability during solidification shrinkage is poor, making it unable to effectively compensate for areas that are about to shrink. It also increases the tendency for premature solidification of the melt in the runner and远离 the gate, increasing the flow resistance for feeding and even blocking the flow path, further reducing the feeding capacity and leading to pore formation. This sand casting defect is exacerbated by low pouring temperatures that restrict the feeding range of the gating system.
When the pouring temperature is increased to 1375–1400 ℃, at shorter pouring times (25–30 s), the volume of mesh elements with shrinkage defects decreases to about 20 cm³, and the average shrinkage porosity reduces to approximately 1.30%, indicating an improvement in shrinkage defects. However, when the pouring time increases to 35 s, the defective mesh element volume reaches 24 cm³, and the shrinkage porosity is 2.06%. This suggests that as the pouring temperature rises, the fluidity of the melt improves, and the feeding capacity is enhanced, thereby reducing shrinkage defects. However, when the pouring time is too long, the filling speed is slower, causing the temperature of the already filled melt to drop rapidly. Consequently, the metal temperature in certain areas becomes too low, solidification occurs prematurely, and the areas about to shrink cannot be replenished in time, resulting in shrinkage pores. The interaction between pouring temperature and pouring time is a key factor in controlling this type of sand casting defect.
When the pouring temperature is further increased to 1425 ℃, the influence of pouring time on casting quality becomes more significant. At a shorter pouring time of 25–30 s, the volume of mesh elements with shrinkage defects is about 18 cm³, and the shrinkage porosity is only 0.86%, indicating minimal shrinkage defects on the casting. However, when the pouring time increases to 35 s, both the defective mesh element volume and shrinkage porosity increase, reaching 25 cm³ and 3.20%, respectively.
| Pouring Temperature (℃) | Pouring Time (s) | Defective Volume (cm³) | Shrinkage Porosity (%) |
|---|---|---|---|
| 1350 | 25 | 26.8 | 2.19 |
| 1350 | 30 | 30.2 | 3.56 |
| 1350 | 35 | 33.5 | 4.17 |
| 1375 | 25 | 22.4 | 1.45 |
| 1375 | 30 | 24.1 | 1.82 |
| 1375 | 35 | 28.3 | 2.71 |
| 1400 | 25 | 19.6 | 1.12 |
| 1400 | 30 | 21.2 | 1.38 |
| 1400 | 35 | 24.0 | 2.06 |
| 1425 | 25 | 18.1 | 0.92 |
| 1425 | 30 | 17.8 | 0.86 |
| 1425 | 35 | 25.0 | 3.20 |
During the casting process, if the filling speed is too fast, the temperature field distribution of the casting becomes uneven, which is not conducive to forming a temperature gradient from the workpiece to the riser and from the workpiece to the gating system, making it difficult to effectively feed the workpiece during solidification. In addition, a fast filling speed may increase the tendency for gas entrapment and inclusions, ultimately affecting the forming quality. Therefore, considering the combined effects of pouring temperature and pouring time on the forming quality, the optimal pouring temperature and pouring time for the MGCI brake drum are 1425 ℃ and 30 s, respectively. At these parameters, the volume of mesh elements with shrinkage defects is minimized, and the shrinkage porosity is reduced to 0.86%, representing a significant reduction in this sand casting defect compared to the original process. The optimization demonstrates that careful control of both thermal and fluid dynamic conditions is essential to produce sound castings free from shrinkage-related sand casting defects.
The relationship between the shrinkage porosity and the process parameters can be described by the following empirical model derived from the full factorial simulation results:
$$
P(T, t) = a_0 + a_1 T + a_2 t + a_3 T^2 + a_4 t^2 + a_5 T t
$$
where P is the shrinkage porosity (%), T is the pouring temperature (℃), t is the pouring time (s), and a₀–a₅ are regression coefficients. The optimal condition (T = 1425 ℃, t = 30 s) yields the minimum porosity of 0.86%, which is a substantial improvement over the original condition (T = 1350 ℃, t = 30 s) that gave a porosity of 3.56%. This quantitative analysis confirms that the sand casting defect can be effectively controlled by optimizing the casting process parameters.
| Metric | Original Process | Optimized Process | Improvement |
|---|---|---|---|
| Pouring Temperature (℃) | 1350 | 1425 | +75 |
| Pouring Time (s) | 30 | 30 | — |
| Defective Volume (cm³) | 30.2 | 17.8 | −41.1% |
| Shrinkage Porosity (%) | 3.56 | 0.86 | −75.8% |
| Defect Location | Ribs near ingate | Minimal defect | — |
The substantial reduction in shrinkage porosity demonstrates that the sand casting defect which was present in the original design has been effectively eliminated through the combined optimization of the gating system structure and the casting process parameters. The improved feeding design ensures that the solidification sequence progresses from the thinner sections toward the thicker sections and the gating system, allowing adequate liquid metal to compensate for volumetric shrinkage during solidification. This systematic approach to sand casting defect prediction and control provides a robust framework for producing high-quality MGCI brake drums.
Experimental Verification of Optimized Process
Based on the optimized gating system structure and casting process parameters, we conducted another casting trial for the MGCI brake drum. The tapping temperature was 1450 ℃, the pouring temperature was 1425 ℃, and the pouring time was 30 s. The cast workpiece shows no visible pores on the appearance or cross-section. Ultrasonic inspection confirmed the absence of internal shrinkage defects. The machined brake drum workpiece exhibits dense internal structure and good surface quality. This experimental verification confirms that the sand casting defect prediction and optimization methodology developed in this study is effective for producing sound castings. By addressing the root causes of the sand casting defect through simulation-guided process design, we have successfully eliminated the shrinkage porosity that was present in the original castings and achieved consistent casting quality.
Conclusions
In this study, we combined modified material property calculations, numerical simulation, and casting experiments to predict and control sand casting defects in modified gray cast iron brake drums. The following conclusions can be drawn from this work:
(1) The solidus and liquidus temperatures of the modified gray cast iron are both reduced compared to HT250, and the variation of solid fraction with temperature is more gradual within the solid–liquid coexistence range. These changes in material properties affect the solidification behavior and the tendency for sand casting defect formation.
(2) The established finite element model for the MGCI brake drum casting can accurately analyze the casting process and predict sand casting defects. The simulation results are consistent with experimental observations, validating the reliability of the model for process optimization.
(3) The MGCI brake drum exhibits good filling behavior, but the rib area, due to its特殊的 heat dissipation location, is prone to forming isolated liquid phase regions during solidification, leading to shrinkage defects. Adjusting the gating system structure to enhance its feeding capacity is essential for mitigating this sand casting defect.
(4) The optimal pouring temperature and pouring time for the MGCI brake drum are 1425 ℃ and 30 s, respectively. Under these parameters, the shrinkage porosity of the workpiece is reduced to 0.86%, representing a 75.8% reduction compared to the original process. The cast brake drums are internally dense, have good surface finish, and are free from sand casting defect.
(5) The integrated approach of material property computation, finite element simulation, and experimental validation provides an effective methodology for predicting and controlling sand casting defects in complex castings, enabling process optimization and quality improvement in production.