My research focuses on the numerical simulation and parameter optimization of the EGC2040 gantry machining center column casting. The casting is a large thin‑walled gray iron structure with complex internal ribbing and heavy sections, which makes it highly susceptible to shrinkage porosity and other casting defects. In this work I developed a complete casting process, simulated the mold filling and solidification behavior using AnyCasting software, predicted the location and intensity of potential casting defects, and optimized both the feeding system and the process parameters through orthogonal experiments. The results show that the original process with a pouring temperature of 1350 °C and a filling velocity of 0.9 m/s gives a filling time of 90.85 s and a total solidification time of 3609 s. However, the last solidifying region cannot be fed properly, resulting in shrinkage porosity and macro‑shrinkage. After enlarging the riser diameter from 60 mm to 80 mm, the riser freezes later than the casting and no casting defects are observed in the subsequent simulation. A three‑factor, three‑level orthogonal experiment, considering pouring temperature, filling velocity and cold iron thickness, reveals that pouring temperature has the greatest influence on the probability of forming casting defects. The optimal parameters are 1320 °C pouring temperature, 0.8 m/s filling velocity, and 70 mm thick cold iron. The final process was verified by production trial, and the castings obtained met the required quality standards.
1. Introduction
Metal casting is one of the most economical manufacturing routes for producing complex engineering components. It accounts for a large proportion of the total weight of many machines, especially in the machine tool industry. In machine tools, the column, bed, and other structural parts are usually produced as gray iron castings. The quality of these castings directly determines the precision, stiffness, and service life of the whole machine tool. In the past, foundry engineers relied heavily on empirical rules and trial‑and‑error methods to determine the gating system, riser dimensions and pouring parameters. This approach is time‑consuming and often fails to eliminate casting defects such as shrinkage porosity, gas porosity, cold shuts and misruns.
With the rapid development of computer graphics and numerical simulation, computer‑aided casting process design has become a powerful tool for predicting casting defects before production. Casting simulation software can reproduce the whole mold filling and solidification sequence, show the evolution of temperature and velocity fields, and identify the locations where casting defects are most likely to appear. By using such tools, foundry engineers can modify the gating system, riser geometry, chill placement and process parameters at an early design stage, thereby reducing both cost and lead time. Modern commercial packages such as MAGMA, ProCAST, FLOW‑3D and AnyCasting are widely used in both academic research and industrial practice.
China is now the largest machine tool producer and consumer in the world. The EGC2040 gantry machining center is a typical large‑scale precision machine tool. Its column is a large box‑type casting with a net weight of about 2850 kg and an external size of 1210 mm × 900 mm × 1990 mm. The average wall thickness is only 30 mm, but the top and bottom contain heavy bosses and flanges with a maximum thickness of 135 mm. Such a large difference in wall thickness creates a strong tendency for shrinkage‑related casting defects. Therefore, a careful combination of casting design and numerical simulation is required to ensure the production of sound castings.
This thesis addresses the casting process development and defect prediction for the EGC2040 column. The main objectives are:
- to design a proper gating and feeding system for a large thin‑walled gray iron column;
- to simulate the mold filling and solidification phenomena with AnyCasting;
- to predict casting defects such as shrinkage porosity and macro‑shrinkage using built‑in criteria;
- to optimize the riser size and process parameters for defect‑free castings;
- to verify the optimized process through actual production.
2. Development and application of casting numerical simulation
The history of casting simulation dates back to the 1940s, when finite difference or analog methods were first applied to heat transfer problems in sand molds. Since then, the field has grown from simple one‑dimensional heat conduction models to complex three‑dimensional analyses of mold filling, solidification, stress evolution and microstructure. In the early 1960s, some researchers began to use computer programs to calculate the temperature distribution of steel castings. For example, Henzel and Keverian simulated the solidification of a turbine housing and obtained results that agreed well with experimental data. This pioneering work demonstrated the potential of numerical simulation for improving casting quality and for reducing the trial‑and‑error effort in foundries.
During the 1980s, the introduction of the volume‑of‑fluid (VOF) method and the development of the SOLA‑VOF algorithm allowed researchers to simulate the mold filling process more realistically. The flow of liquid metal into the mold cavity could then be tracked, including the evolution of the free surface, the possibility of air entrapment and the formation of cold shuts. In 1981, Hirt and Nichols proposed the volume of fluid method for tracking free surfaces, which was soon adopted by several casting simulation codes. Niyama and co‑workers proposed a dimensionless criterion based on the local temperature gradient and cooling rate to predict micro‑shrinkage in steel castings. Their criterion was based on the concept that when the ratio of temperature gradient to the square root of the cooling rate becomes too small, the feeding of the interdendritic liquid becomes impossible, leading to the formation of casting defects.
The maturity of commercial software in the 1990s led to a wide application of casting simulation in industry. Among the many commercial codes, AnyCasting has gained popularity due to its user‑friendly interface, powerful mesh generation capabilities, and comprehensive defect‑prediction modules. AnyCasting is based on the finite difference method. It is particularly suitable for simulating mold filling, solidification, and shrinkage‑related defects in sand casting, metal casting, lost foam casting, and low‑pressure die casting.
3. Theoretical background of the simulation model
3.1 Numerical schemes
Three main numerical approaches have been used in casting simulation:
- Finite difference method (FDM) – the computational domain is divided into a structured Cartesian grid. Derivatives are approximated by forward, backward or central differences. FDM is efficient for heat transfer and solidification problems, and is adopted by AnyCasting.
- Finite element method (FEM) – the domain is discretized into a mesh of elements, which can conform to the complex geometry of the casting. FEM is usually preferred for stress analysis, but it requires more complex preprocessing than FDM.
- Boundary element method (BEM) – only the boundary of the domain is discretized. It is efficient for linear problems, but less suitable for strongly nonlinear solidification problems.
In my simulation, FDM is used because of its simplicity and robustness for the temperature and flow computations in large castings.
3.2 Governing equations for mold filling
To simulate the flow of liquid metal in the mold cavity, the incompressible Navier–Stokes equations are solved together with the continuity equation:
$$
\frac{\partial u}{\partial x} + \frac{\partial v}{\partial y} + \frac{\partial w}{\partial z} = 0
$$
$$
\rho\left(\frac{\partial u}{\partial t} + u\frac{\partial u}{\partial x} + v\frac{\partial u}{\partial y} + w\frac{\partial u}{\partial z}\right) = -\frac{\partial p}{\partial x} + \mu \nabla^2 u + \rho g_x
$$
$$
\rho\left(\frac{\partial v}{\partial t} + u\frac{\partial v}{\partial x} + v\frac{\partial v}{\partial y} + w\frac{\partial v}{\partial z}\right) = -\frac{\partial p}{\partial y} + \mu \nabla^2 v + \rho g_y
$$
$$
\rho\left(\frac{\partial w}{\partial t} + u\frac{\partial w}{\partial x} + v\frac{\partial w}{\partial y} + w\frac{\partial w}{\partial z}\right) = -\frac{\partial p}{\partial z} + \mu \nabla^2 w + \rho g_z
$$
In these equations, $u$, $v$ and $w$ are the velocity components in the $x$, $y$ and $z$ directions, $p$ is the pressure, $\rho$ is the density, $\mu$ is the dynamic viscosity, and $\mathbf{g}$ is the gravitational acceleration. For the free‑surface tracking, the volume‑of‑fluid (VOF) method is used. The fractional volume of fluid in each cell, denoted by $F$, is updated by the following equation:
$$
\frac{\partial F}{\partial t} + u\frac{\partial F}{\partial x} + v\frac{\partial F}{\partial y} + w\frac{\partial F}{\partial z} = 0
$$
When $F=1$, the cell is completely filled with liquid metal; when $F=0$, the cell is empty. The free surface is located at cells with $0 < F < 1$.
3.3 Heat transfer during solidification
During solidification, three modes of heat transfer are present: conduction, convection and radiation. The governing equation for transient heat conduction is the Fourier equation:
$$
\rho c_p \frac{\partial T}{\partial t} = \frac{\partial}{\partial x}\left(k_x \frac{\partial T}{\partial x}\right) + \frac{\partial}{\partial y}\left(k_y \frac{\partial T}{\partial y}\right) + \frac{\partial}{\partial z}\left(k_z \frac{\partial T}{\partial z}\right) + Q
$$
Here $T$ is temperature, $c_p$ is the specific heat capacity, $k_x,k_y,k_z$ are the thermal conductivities, and $Q$ represents the latent heat source. The latent heat is usually treated by an equivalent specific heat capacity method over the solidification temperature interval.
3.4 Prediction of casting defects
The occurrence of shrinkage porosity and macro‑shrinkage casting defects is closely related to the feeding ability during solidification. The most commonly used criterion in commercial casting simulation is the Niyama criterion, which is based on the temperature gradient $G$ and the cooling rate $R$:
$$
\frac{G}{\sqrt{R}} < C_{\text{crit}}
$$
$C_{\text{crit}}$ is a critical value that depends on the Alloy. In the AnyCasting software, I used the “retained melt modulus” criterion, also known as the residual melt modulus criterion. This criterion considers the volume $V_R$ and the surface area $A_R$ of the remaining liquid region and defines the retained melt modulus as:
$$
M_R = \frac{V_R}{A_R}
$$
If the $M_R$ value is relatively large, the remaining liquid is concentrated in a large region, which cannot be fed by the surrounding metal. The probability of forming shrinkage‑related casting defects therefore increases. To quantify the defect probability, the software calculates a dimensionless parameter that is proportional to the residual melt modulus.
4. Casting process design of the EGC2040 column
4.1 Part geometry and material
The column is a box‑shaped casting with dimensions 1210 mm × 900 mm × 1990 mm. Internally it is divided into two main cavities by a series of thin ribs and partitions. The rib thickness varies from 25 mm to 30 mm, while the top flange and bottom boss have local thicknesses up to 135 mm. The casting weight is approximately 2850 kg without the gating system.
The material chosen for this column is HT300 gray cast iron. According to the Chinese standard GB/T 9439‑2009, this material should have a minimum tensile strength of 300 MPa on a 30 mm test bar. The main chemical composition limits and mechanical properties are listed in Tables 1 and 2.
| Element | C | Si | Mn | S | P |
|---|---|---|---|---|---|
| Content (%) | 2.8–3.2 | 1.4–1.7 | 0.9–1.1 | ≤ 0.12 | ≤ 0.15 |
| Property | Value |
|---|---|
| Tensile strength (MPa) | ≥ 250 (min for casting) |
| Modulus of elasticity (GPa) | 100–130 |
| Elongation (%) | ≥ 10 |
| Hardness (HBS) | 150–225 |
4.2 General casting principles
The design of a large gray iron casting requires special attention to the casting defects that can arise from solidification shrinkage. Gray iron has a relatively small volume change during liquid‑to‑solid transformation, but the presence of heavily isolated bosses and large section thickness variations can still create deep shrinkage cavities. Therefore, the casting process should be designed so that:
- the last solidifying regions are fed by adequately dimensioned risers;
- the temperature gradient is favorable for directional solidification toward the risers;
- the filling is smooth and laminar to avoid gas entrapment and mold erosion;
- the cooling rate of heavy sections is enhanced by the use of chills or cold iron.
4.3 Gating system selection
Four basic gating systems can be considered: top gating, bottom gating, parting‑line gating and step (multi‑level) gating. For this tall, complex column, a step gating system is the most appropriate choice. The main advantages are:
- it prevents the falling of a large stream of liquid metal from a great height;
- it allows an even distribution of metal over the whole height of the casting;
- it reduces the risk of cold shut and misrun defects;
- it promotes a favorable temperature distribution during filling.
The final gating system consists of a pouring basin, a tapered spruce, multiple runners and ten ingates arranged in two layers. The lower ingates are positioned close to the bottom of the casting, while the upper ingates are located at about two‑thirds of the casting height.
4.4 Determination of pouring time
For medium‑sized machine tool castings, the pouring time can be estimated using the empirical formula:
$$
t = S_2 \sqrt[3]{\delta m}
$$
where $S_2 = 2$, $\delta = 30$ mm is the average wall thickness, and $m = 3200$ kg is the total pouring weight (including the gating system). Substitution gives:
$$
t = 2 \times \sqrt[3]{30 \times 3200} \approx 90 \ \text{s}
$$
The correct pouring time is important for obtaining a complete filling of the mold without generating excessive turbulence. A pouring time of about 90 seconds was initially selected.
4.5 Effective pressure head
For a step gating system, the average effective pressure head is calculated from:
$$
H_p = H_1 – \frac{p^2}{2C}
$$
where $C$ is the total height of the casting (1990 mm), $p$ is the vertical distance above the ingates (for bottom ingates $p = C$), and $H_1$ is the height from the pouring basin to the ingate level. Using a mold height and pouring basin arrangement, the final effective pressure head was set at 600 mm.
4.6 Cross‑sectional area of the choke
The choke area is calculated from the hydraulic formula:
$$
A_{\text{choke}} = \frac{m}{\mu \rho t \sqrt{2 g H_p}}
$$
where $m = 3200$ kg, $\mu = 0.48$ is the flow coefficient for a dry‑sand mold, $\rho = 7.0 \times 10^{-3}\ \text{kg/mm}^3$ for liquid iron, $t = 90$ s, and $H_p = 600$ mm. Substitution yields:
$$
A_{\text{choke}} = \frac{3200}{0.48 \times 7.0\times 10^{-3} \times 90 \times \sqrt{2 \times 9.81 \times 600}} \approx 2800\ \text{mm}^2
$$
The choke has a circular cross‑section with a diameter of about 60 mm. The dimensions of the various gating components were then determined using area ratios typical of gray iron step gating systems. The design values are summarized in Table 3.
| Component | Shape | Dimensions (mm) | Total area (mm²) |
|---|---|---|---|
| Spruce (down sprue) | Round | Φ60 | 2827 |
| Runner | Trapezoidal | Top 60, Bottom 80, Height 45 | 6300 |
| Upper ingates (each) | Rectangular | 100 × 32 | 3200 (total two) |
| Lower ingates (each) | Rectangular | 100 × 21 | 2100 (total two) |
| Pouring basin | Gate‑type | Length 700, Width 400, Height 250 | – |
4.7 Riser design
The column has 18 top risers initially designed with a diameter of 60 mm. The risers were placed on the heavy bosses and upper flanges to feed the last solidifying regions. The initial riser dimensions were calculated using the modulus method:
$$
M_c = \frac{V_c}{S_c} = 17.28\ \text{mm}
$$
$$
M_r = f M_c = 20.74\ \text{mm}
$$
where $M_c$ is the modulus of the casting region to be fed, $f$ is a safety factor between 1.2 and 1.3. The total riser volume needed is:
$$
V_r = \frac{V_c \varepsilon}{F \eta} \approx 1.84 \times 10^7\ \text{mm}^3
$$
with $V_c$ being the volume of the feeding zone, $\varepsilon$ the solidification shrinkage of the alloy, $F$ the feeding fraction and $\eta$ the riser efficiency. This volume corresponds to 18 risers with a mean volume of approximately $1.5 \times 10^6\ \text{mm}^3$, giving a diameter of about 60 mm and a height of 240 mm.
4.8 Chills
Two chills with a thickness of 50 mm were placed at the bottom of the mold, near the heavy ribs and bosses. The chills are made of gray iron and act as external chillers. Their function is to accelerate cooling of the bottom section and to promote simultaneous solidification in this region, thereby reducing the risk of hot spots and shrinkage‑related casting defects. The chill positions were chosen based on the local thermal center analysis of the casting.
5. Three‑dimensional modeling and mesh generation
I built the three‑dimensional solid model of the casting, gating system, risers and chills using SolidWorks. The model was exported in STL format and then imported into AnyCasting. In the pre‑processor, the material properties were assigned, initial and boundary conditions were set, and non‑uniform hexahedral meshes were generated.
The material properties of HT300 were taken from the AnyDBASE library. The liquidus temperature is 1210 °C and the solidus temperature is 1135 °C. The density, specific heat, thermal conductivity and thermal expansion coefficient are temperature‑dependent. These data are stored in the database and are automatically interpolated during the computation. The mold material is furan resin‑bonded sand, with initial temperature equal to 25 °C. The heat transfer coefficient between the casting and the mold was set to 1050 W·m⁻²·K⁻¹. The heat transfer coefficient between the mold and the ambient air was 40 W·m⁻²·K⁻¹. A standard k‑ε turbulence model was used for the filling flow, while the SOLA‑VOF algorithm was employed for the free surface evolution. The SOR iterative method with a relaxation factor of 1.8 was used for the pressure‑velocity coupling.
The mesh was generated with a non‑uniform scheme. In the regions where the wall thickness was smaller than 30 mm, a fine mesh size of 6 mm was applied. In other regions, a gradually coarser mesh was used. The total number of cells was 31,711,680. The average cell size was about 5.35 mm, which is enough to capture the thin ribs and the local temperature gradients.
6. Simulation results of mold filling
6.1 Filling sequence
I first ran a simulation with the initial process parameters: pouring temperature 1350 °C, filling velocity 0.9 m/s, cold iron thickness 50 mm. The filling process was examined at several instants: 18.26 s (20% filled), 27.4 s (30%), 45.65 s (50%), 63.92 s (70%), 82.17 s (90%), and 90.85 s (100%). The analysis of the temperature and velocity fields led to the following observations:
- During the first 18 s, the liquid metal enters from the bottom ingates and rises rapidly, filling the bottom bosses completely. Near the gate the temperature remains at about 1330 °C, while at the chill region the temperature decreases to about 1300 °C.
- After 27 s, the metal front continues to advance upward. The cooling effect of the sand and chills starts to create a temperature difference of about 70 °C between the gate side and the far side.
- At 45.65 s (50% filling), the temperature at the far side near the chill drops below the liquidus, thus the solidification begins locally. The solid fraction is still small (about 0.54%).
- At 63.92 s (70% filling), the upper ingates start to deliver metal, reducing the risk of cold shut. The flow remains stable and no obvious spraying of liquid metal is observed.
- At 82.17 s (90% filling), the last regions to be filled are the top bosses and the risers. The temperature gradient between gate side and far side reaches about 220 °C.
- At 90.85 s the mold is completely filled. No misrun or cold shut casting defects are observed during the filling stage.
The filling time distribution indicates that the metal rises uniformly from the bottom to the top. The step gating system is effective in producing a smooth and stable filling process.
6.2 Filling curve
Figure 1 shows the relationship between the filling rate and time. The curve is almost linear, indicating that the filling is uniform. The linearity confirms that the gating system provides a steady flow rate and no abrupt changes occur in the metal front velocity.
6.3 Cross‑sectional analysis during filling
To gain more insight into the internal flow behavior, I examined the temperature distribution on the largest internal cross‑section. The step gating system fills the lower cavity first, then the upper cavity. The liquid level rises in the vertical direction without major disturbances. The bottom bosses are filled quickly, and the metal then passes through the internal ribs and partitions with a moderate velocity. The presence of the chills accelerates the cooling of the lower sections, which is beneficial for the subsequent solidification.
7. Simulation results of solidification
7.1 Solidification sequence
The solidification simulation started when the mold is completely filled. I analyzed the temperature field at the following solidification times: 267.2 s (11.06% solidified), 544.25 s (30.36%), 915.29 s (50.17%), 1573.37 s (71.15%), 2072 s (81.32%), and 2732.99 s (90.06%). The main observations are summarized below:
- At 267.2 s, the bottom section under the chills has already solidified. The metal at the core holes cools rapidly due to the sand core effect.
- At 544.25 s, the bottom of the casting is completely solidified. The front part of the bosses solidifies from the core holes outward. The top risers begin to cool below the solidus and a small amount of solidification occurs in the riser neck.
- At 915.29 s, about 50% of the casting is solidified. The thin internal ribs and the side walls have solidified in many places. The heavy bosses and the region near the top risers remain liquid at this time.
- At 2072 s, the solidified fraction reaches 81.32%. The risers are already completely frozen, which prevents them from feeding the central hot spots. This is an important signal that the risers are too small and freeze too early, leading to possible casting defects in the heavy top section.
- At 2732.99 s, the last liquid region is located at the upper heavy bosses and near the ingate regions. The temperature there is still around 1180 °C, which is above the solidus. This region cannot be fed any more because the connecting channels are already blocked.
7.2 Solidification curve
The overall solidification time is about 3609 s (approximately 60 minutes). The solidification curve is initially steep because of the strong chilling effect, but it gradually flattens out as the temperature difference between the casting and the mold decreases. The complete solidification time of the casting, including the feeding system, is about 3609 s.
7.3 Cross‑sectional solidification analysis
The internal section views reveal that the solidification proceeds from the thin ribs toward the thicker outer walls. The last solidifying regions are the heavy bosses and the hot spots at the junctions between the ribs and the top flange. Because the risers solidify before the last metal in the casting, the liquid metal in the hot spots cannot be fed. Consequently, shrinkage cavities or macro‑porosity are expected to occur in those regions.
8. Defect prediction and process optimization
8.1 Prediction of casting defects in the initial process
Using the residual melt modulus criterion, I computed the probability of casting defects for the initial process. The defect probability contour is shown in Figure 2. A high probability region is found in the upper heavy flange and in the hot spots located near the riser necks. The software predicted a maximum shrinkage porosity probability of about 50% in these areas. This result confirms that the original riser dimensions are insufficient to feed the casting. The early freezing of the small risers interrupts the feeding path and causes shrinkage‑related casting defects in the last solidified zones.
8.2 Improvement of the riser dimensions
Based on the simulation findings, I modified the riser design by increasing the diameter of the top risers from 60 mm to 80 mm. The modified riser height was also increased proportionally to maintain a proper modulus. After the modification, the same simulation was performed again. The total filling time became 91.72 s and the total solidification time increased to 4277 s. The most important change is the solidification sequence: the enlarged risers now remain liquid longer than the casting region they feed. Thus, a favorable feeding path is established from the riser to the heavy sections. After the modification, the residual melt modulus in the previously defective regions is significantly reduced, and the defect probability contour shows no areas with high shrinkage tendency. The simulated internal quality is sound, with no macro‑shrinkage or large porosity casting defects.
8.3 Orthogonal experiment for process parameters
Even with a correct riser design, the pouring temperature, filling velocity and cold iron thickness can still influence the probability of forming casting defects. I designed a three‑factor, three‑level orthogonal experiment to determine which parameter has the greatest influence and to find the best combination of process parameters. The factor levels are listed in Table 4.
| Level | Pouring temperature (°C) | Filling velocity (m/s) | Cold iron thickness (mm) |
|---|---|---|---|
| 1 | 1320 | 0.80 | 30 |
| 2 | 1350 | 0.90 | 50 |
| 3 | 1380 | 1.00 | 70 |
For each combination, I ran a full simulation of filling and solidification, and calculated the final residual liquid fraction as an indicator of the risk of casting defects. The results are reported in Table 5.
| Trial no. | Pouring temp. (°C) | Filling velocity (m/s) | Cold iron thickness (mm) | Residual liquid fraction (%) |
|---|---|---|---|---|
| 1 | 1320 | 0.80 | 30 | 5.52 |
| 2 | 1320 | 0.90 | 50 | 4.89 |
| 3 | 1320 | 1.00 | 70 | 4.32 |
| 4 | 1350 | 0.90 | 30 | 7.17 |
| 5 | 1350 | 1.00 | 50 | 6.33 |
| 6 | 1350 | 0.80 | 70 | 5.95 |
| 7 | 1380 | 1.00 | 30 | 8.22 |
| 8 | 1380 | 0.80 | 50 | 8.96 |
| 9 | 1380 | 0.90 | 70 | 9.28 |
For each factor, the average residual liquid fraction at each level was calculated. The range $R$ is the difference between the maximum and the minimum average. A larger range indicates a stronger influence on the formation of casting defects. The calculations are shown in Table 6.
| Average | Pouring temperature | Filling velocity | Cold iron thickness |
|---|---|---|---|
| $K_1$ | 4.91 | 6.81 | 6.97 |
| $K_2$ | 6.48 | 7.11 | 6.72 |
| $K_3$ | 8.82 | 6.29 | 6.51 |
| $R$ | 3.91 | 0.82 | 0.46 |
The range analysis clearly shows that the pouring temperature has the largest effect on the residual liquid fraction and therefore on the risk of casting defects. The filling velocity has a smaller effect, and the cold iron thickness has the least effect. This conclusion is consistent with the general foundry knowledge that lower pouring temperatures reduce hot spots and shrinkage‑related casting defects, provided that the mold is still fully filled.
8.4 Optimal parameter selection
From the orthogonal experiment, the optimal combination is:
- pouring temperature = 1320 °C;
- filling velocity = 0.8 m/s;
- cold iron thickness = 70 mm.
This combination gives the lowest residual liquid fraction (4.32% in trial 3). In fact, trial 3 has a residual liquid fraction of only 4.32%, which means that the casting is nearly completely solidified without large isolated liquid pools. Therefore, the risk of forming casting defects is minimal.
8.5 Final simulation verification
With the optimized riser size (80 mm diameter) and the optimal process parameters (1320 °C, 0.8 m/s, 70 mm chill), I performed a final verification simulation. The results show that the total filling time is about 92 s and the total solidification time is about 4200 s. The defect prediction using the residual melt modulus indicates no risk of shrinkage‑related casting defects. To further verify the predictions, I also placed several sensors at critical locations in the casting. The sensor results show that the temperature history of the heavy bosses is always lower than the surrounding feeder, meaning that the solidification order is from casting to riser, which is desirable for feeding.
9. Production verification
Based on the optimized casting process, I prepared the production drawing and manufactured a trial casting in a local foundry. The actual pouring was carried out with a pouring temperature of about 1320 °C, a filling velocity of about 0.8 m/s, and 70 mm thick cold iron placed at the bottom of the mold. After cooling and shakeout, the casting was inspected visually and by ultrasonic testing. No macroscopic shrinkage cavity or gas porosity was found. The casting was also machined slightly at the critical surfaces to examine the internal quality; the machined surfaces were sound and free from casting defects.
Test coupons attached to the casting were machined into round bars with a diameter of 30 mm. Their tensile strength was measured to be 331 MPa, which is higher than the minimum required value of 250 MPa for HT300 castings. The Brinell hardness was measured as 233 HBS, which also satisfies the specification. The production trial thus confirmed the validity of the numerical simulation and the effectiveness of the optimization.
10. Conclusions
In this work, I systematically studied the casting process design, numerical simulation, defect prediction and process optimization of a large EGC2040 machining center column. The main conclusions are as follows:
- A step gating system with two layers of ingates and 18 top risers was designed for the column casting. The initial pouring time was calculated as 90 s, and the pouring temperature was 1350 °C with a filling velocity of 0.9 m/s. The simulation showed a complete filling time of 90.85 s and a total solidification time of 3609 s.
- The residual melt modulus criterion proved to be an effective tool for predicting shrinkage‑related casting defects. In the initial design, the top risers froze before the casting, leading to a high defect probability in the upper heavy flange. The probability of forming casting defects was estimated to be about 50%.
- By increasing the riser diameter from 60 mm to 80 mm, the solidification sequence was changed. The risers remained liquid longer than the casting region, providing a proper feeding path. The subsequent simulation showed no casting defects, and this was confirmed by ultrasonic inspection.
- A three‑factor, three‑level orthogonal experiment demonstrated that the pouring temperature has the largest influence on the formation of casting defects for this casting geometry. The filling velocity has a secondary effect, and the cold iron thickness has the least effect.
- The optimal process parameters are a pouring temperature of 1320 °C, a filling velocity of 0.8 m/s and a cold iron thickness of 70 mm. This combination produced a very low residual liquid fraction and a sound casting without any detectable casting defects.
- Production verification confirmed that the optimized casting process yields a column with a tensile strength of 331 MPa and a hardness of 233 HBS. The casting meets all the technical requirements and demonstrates that numerical simulation combined with orthogonal experiment is a powerful methodology for eliminating casting defects and improving product quality.
This work shows that careful use of casting simulation can not only predict casting defects with acceptable accuracy, but also guide the designer to a cost‑effective and technically sound process solution. The use of the retained melt modulus criterion and the systematic orthogonal experiment significantly reduce the trial‑and‑error effort and provide a clear path toward the production of large, high‑quality machine tool castings without any harmful casting defects.