In the foundry industry, ensuring the integrity of cast components while minimizing production costs and lead times remains a persistent challenge. As a researcher working on the interface between traditional casting technology and modern computational tools, I have dedicated my work to the application of casting simulation technology, commonly referred to as casting CAE, to study and improve the manufacturing of complex valve body castings. The central motivation is to systematically predict and eliminate casting defects such as shrinkage porosity, gas entrapment, and misruns, which have long plagued foundries. Throughout this article, I will share my findings from a series of investigations into valve body castings, leveraging three-dimensional modeling, numerical simulation of mold filling and solidification, and rigorous comparison with actual production results.
Valve body castings are among the most challenging components to produce due to their intricate geometries, thick-walled sections, and demanding service conditions. They are often used in high-pressure and high-temperature environments, where even microscopic internal porosity can lead to catastrophic leakage. Traditional trial-and-error methods for designing gating and risering systems consume enormous resources and still leave significant uncertainty regarding final internal soundness. With the advent of computer-based casting simulation, it has become possible to visualize the entire casting process before metal is poured. This allows engineers to identify critical regions where casting defects are likely to form and to modify the feeding system, chills, and pouring parameters accordingly.
In this article, I will describe how I combined a commercial CAD package with a robust casting simulation solver to analyze and optimize a range of valve body components, including heavy-walled gate valves and flanged fittings. The results demonstrate that systematic use of casting CAE not only reduces the incidence of casting defects but also improves process yield and shortens product development cycles.
1. Introduction and Significance of Casting Simulations
Castings represent a fundamental manufacturing route for producing metal components with complex shapes. In the global economy, castings are indispensable for industries such as energy, transportation, and heavy machinery. However, the casting process is inherently prone to various discontinuities. The most common casting defects include macro-shrinkage, micro-shrinkage or dispersed porosity, hot tearing, cold shuts, gas holes, sand inclusions, and oxide films. These casting defects not only degrade mechanical properties but also cause leaking in pressure-containing parts. The ability to predict and mitigate such defects before committing to production tooling is enormously valuable.
Numerical simulation of mold filling and solidification has matured over the past few decades. Early work in the 1960s laid the foundation for heat transfer calculations in castings. Subsequent advances introduced computational fluid dynamics for free-surface flows, and later coupling of thermal, flow, and stress fields. Today, commercial simulation tools are available that can model the complete casting process, including turbulent flow, heat conduction, phase transformation, and the formation of microporosity. In my research, I utilized one such platform, which I will refer to as the casting CAE system, to analyze valve body castings. This tool enables engineers to set up a virtual experiment, test alternative feeding designs, and obtain quantitative data on the temperature field, velocity field, solid fraction evolution, and defect indicators.
The importance of this approach cannot be overstated. A typical traditional development cycle for a new valve body involves multiple rounds of prototyping, machining, pressure testing, and defect analysis. Each iteration may take weeks and consumes expensive materials and energy. By contrast, a well-calibrated simulation can identify the likely casting defects within hours, allowing the engineer to modify riser sizes, chill placements, or gating dimensions directly on the computer model. The net effect is a substantial reduction in time-to-market and a significant improvement in first-pass yield.
2. Theoretical Basis for Casting Process Design
Before delving into the simulation results, it is essential to review the classical analytical foundations that informed the initial process layouts. These principles are still useful for establishing a reasonable baseline design that can then be refined through numerical analysis. The main components to be designed for a sand casting of steel or ductile iron are the risers, the gating system, and the chills.
2.1 Riser Design Principles
Risers are reservoirs of liquid metal that feed the solidifying casting to compensate for volumetric shrinkage. A sound riser must satisfy three conditions: (1) it must remain liquid longer than the casting section it feeds, (2) it must contain enough liquid metal to supply the solidification shrinkage, and (3) there must be a clear feeding path between the riser and the region being fed. The concept of modulus is commonly used to compare the solidification time of a casting section and a riser. The modulus \(M\) is defined as:
$$M = \frac{V}{A}$$
where \(V\) is the volume and \(A\) is the cooling surface area. For a given alloy and mold, the solidification time is proportional to \(M^2\). For a steel casting, a typical rule is that the riser modulus must be at least \(1.2\) times the modulus of the casting section to ensure that the riser freezes last. Additionally, the riser volume must be sufficient to compensate for the total liquid contraction and solidification contraction. The solidification shrinkage of steel is approximately 5% to 6% by volume. The riser efficiency \(\eta\) is defined as the fraction of the riser volume that is available for feeding. Typical values for various riser types are shown in Table 1.
| Riser type | Feeding efficiency \(\eta\) (%) |
|---|---|
| Cylindrical open riser | 14–20 |
| Cylindrical blind riser | 20–30 |
| Spherical riser | 25–35 |
| Insulated riser | 30–45 |
| Exothermic riser | 40–60 |
| Gas-pressurized riser | 35–50 |
To calculate the required riser size, the method of modulus ratio is often employed. For a top riser, the condition is:
$$M_{riser} = 1.2\,M_{casting}$$
For a side riser, a slightly larger ratio may be needed:
$$M_{riser} = 1.3\,M_{casting}$$
Once the modulus requirement is satisfied, the volume is checked by the feeding balance equation:
$$V_{riser} \cdot \eta = V_{casting} \cdot \varepsilon_{shrinkage}$$
where \(V_{riser}\) is the effective riser volume, \(V_{casting}\) is the volume of the casting section to be fed, and \(\varepsilon_{shrinkage}\) is the volumetric solidification shrinkage fraction. In practice, one first determines the modulus-based dimensions and then verifies the volume with the above equation. If the volume is insufficient, the riser height is increased.
2.2 Gating System Design
Proper gating design minimizes turbulence, prevents slag entrapment, and controls the temperature distribution in the mold. The three main types of gating systems based on the pouring position are top, bottom, and stepped. For large steel castings, bottom gating is often preferred because it produces smooth filling and avoids splashing. The cross-sectional areas of the sprue, runner, and ingates follow a ratio that determines the behavior of the liquid metal. A so-called expanding gating system, where \(A_{sprue} < A_{runner} < A_{ingates}\), is recommended for easily oxidized alloys such as steel and aluminum. The minimum cross-sectional area (usually the sprue exit area) can be computed using a hydraulic approximation:
$$A_{min} = \frac{m}{\rho \cdot \mu \cdot t \cdot \sqrt{2g \cdot H_p}}$$
In this formula, \(m\) is the total mass of liquid metal poured, \(\rho\) is the density of the molten metal, \(\mu\) is a flow loss coefficient, \(t\) is the pouring time, \(g\) is the gravitational acceleration, and \(H_p\) is the average effective static pressure head. The flow coefficient \(\mu\) varies from \(0.3\) to \(0.7\) depending on the mold material, gating system complexity, and superheat. For steel castings, \(\mu\) is usually selected between \(0.4\) and \(0.6\). The pouring time \(t\) for heavy steel castings is often determined from empirical charts based on the nominal wall thickness and the mass of the casting.
A simpler alternative, especially for large castings, is the “reverse engineering” approach where the ingate number and size are chosen from experience, and then the sprue and runner dimensions are scaled according to the gating ratio. This approach is frequently used in valve body foundries because of its flexibility and speed.
2.3 Chill Design and Its Influence
Chills are used to accelerate local solidification, thereby promoting directional solidification and eliminating hot spots. External chills are inserted into the mold surface, while internal chills are embedded in the casting itself. The weight of an external chill can be estimated from the heat balance between the casting section and the chill. For a section of volume \(V_1\) that has an adjacent thicker section \(V_2\), the required chill weight \(W_c\) is given by:
$$W_c = \frac{\rho \cdot L \cdot (V_2 – V_1)}{c_c \cdot (T_{solidus} – T_0)}$$
where \(L\) is the latent heat of solidification, \(c_c\) is the specific heat capacity of the chill material (usually steel), \(T_{solidus}\) is the solidus temperature of the casting alloy, and \(T_0\) is the initial temperature of the chill. The denominator may include a factor accounting for the fraction of latent heat absorbed. In my experience, the proper placement of chills can dramatically reduce the number of risers required and improve the internal quality of thick-walled sections.
2.4 Effect of Insulating Riser Sleeves
Insulating sleeves reduce the heat lost from the riser surface, effectively increasing the modulus of the riser. For an insulating sleeve, the apparent modulus \(M_{riser}’\) can be calculated as:
$$M_{riser}’ = \frac{V_{riser}}{A_{bottom} + A_{surface}}$$
where \(A_{bottom}\) is the area in contact with the casting and \(A_{surface}\) is the exposed top area. For a blind insulating riser, the side area is negligible, and the modulus becomes:
$$M_{riser}’ = \frac{V_{riser}}{A_{top}}$$
With insulating sleeves, the riser feeding efficiency can increase substantially, and the riser size can be reduced by 20–30% compared to a conventional sand riser.
3. Mathematical Models for Casting CAE
The simulation tool used in this work is based on the finite difference method (FDM) for solving the governing conservation equations. The comprehensive model couples fluid flow, heat transfer, and solidification. The key equations are described below.
3.1 Energy Transport
For a casting solidifying in a sand mold, the energy equation is typically expressed in terms of enthalpy \(H\) to account for latent heat evolution:
$$\frac{\partial (\rho H)}{\partial t} + \nabla \cdot (\rho \mathbf{u} H) = \nabla \cdot (k \nabla T)$$
where \(\rho\) is the density, \(H\) is the enthalpy, \(\mathbf{u}\) is the velocity vector, \(k\) is the thermal conductivity, and \(T\) is the temperature. The enthalpy method allows a unified treatment of the solid and liquid phases. The latent heat is included in the enthalpy relation:
$$H = \int c_p \, dT + f_l \cdot L$$
where \(c_p\) is the specific heat, \(f_l\) is the liquid fraction, and \(L\) is the latent heat of fusion. The liquid fraction is assumed to vary linearly between the liquidus and solidus temperatures, or more realistically with a lever rule or Scheil model.
3.2 Momentum Transport
For mold filling, the incompressible Navier–Stokes equations are solved. The flow is treated as a laminar or turbulent flow of a Newtonian fluid. The momentum equation in the \(x\)-direction is:
$$\rho \left( \frac{\partial u}{\partial t} + u\frac{\partial u}{\partial x} + v\frac{\partial u}{\partial y} + w\frac{\partial u}{\partial z} \right) = -\frac{\partial p}{\partial x} + \mu_{eff} \nabla^2 u + \rho g_x$$
where \(u, v, w\) are the velocity components, \(p\) is the pressure, \(\mu_{eff}\) is the effective viscosity including turbulence effects, and \(g_x\) is the gravitational acceleration component. The free surface is tracked using the volume-of-fluid (VOF) method or a similar technique. The continuity equation enforces mass conservation:
$$\frac{\partial u}{\partial x} + \frac{\partial v}{\partial y} + \frac{\partial w}{\partial z} = 0$$
During solidification, the flow in the mushy zone is damped by a drag force. The momentum equations are modified by adding a source term that is proportional to the liquid fraction and the velocity:
$$S_{mush} = -K_0 \frac{(1-f_l)^2}{f_l^3 + \epsilon} \mathbf{u}$$
where \(K_0\) is a permeability constant and \(\epsilon\) is a small number to avoid division by zero. This term effectively stops flow when the liquid fraction drops below a critical value.
3.3 Mass Transport
For alloys with significant solute segregation, species conservation may be needed, but in many valve body simulations, only thermal and momentum transport are considered. The mass conservation for the liquid is inherently satisfied by the continuity equation. For shrinkage prediction, the feeding flow is a consequence of the pressure drop in the mushy zone. A common approach is to calculate a temperature gradient and cooling rate based on the Niyama criterion.
3.4 Defect Prediction Criteria
Several post-processing criteria are used to predict casting defects, especially shrinkage porosity. The most widely used is the Niyama criterion, defined as:
$$N = \frac{G}{\sqrt{\dot{T}}}$$
where \(G\) is the thermal gradient and \(\dot{T}\) is the local cooling rate at the time when the liquid fraction reaches a critical value (usually 0.2 or 0.3). A low Niyama value indicates that the liquid cannot be fed through the dendritic network, leading to micro-porosity. In my simulations, I used both the Niyama criterion and a direct feeding resistance criterion. The latter examines the isolated liquid regions after the solid fraction exceeds a critical value. If an isolated liquid pocket has no connection to any riser, it will form a shrinkage cavity upon further cooling.
Another important criterion is the hot spot or thermal modulus. The simulation calculates the time of complete solidification at every location. Regions that solidify later than their surroundings are potential hot spots where casting defects may appear. The combination of these criteria gives a reliable picture of the expected defect distribution.
4. Modeling Workflow and Simulation Setup
My typical workflow begins with the creation of a three-dimensional solid model of the casting, including all attached risers, gating channels, and chills. I used a popular CAD system to generate the geometry. Since the simulation software requires a watertight volume mesh, every separate component—casting, riser, gating system, chill—is exported as an individual part. The parts are then imported into the casting simulation environment, where the material properties, boundary conditions, and numerical parameters are assigned.
The first step in the simulation is mesh generation. The simulation tool uses a finite difference grid, which requires an octree-based mesh. The user specifies a base mesh size and an optional refinement level. For valve bodies with thin flanges and thick walls, I found that a base mesh size of 10 mm, with automatic refinement in regions of small features, provides a good balance between accuracy and computation time. Table 2 lists the mesh settings I typically used.
| Parameter | Value |
|---|---|
| Wall thickness (mm) | 8 |
| Refinement level | 2 |
| Minimum cell size (mm) | 4 |
| Smoothness factor | 2.0 |
| Cell aspect ratio limit | 5 |
After meshing, the materials are assigned. For steel castings, I used the built-in database of low-carbon steel with properties aligned to the German standard GS-C25 (comparable to ASTM WCB). The mold material was sodium silicate sand, which is common in the local foundry. Table 3 shows the thermal-physical data used in the simulations.
| Property | Casting steel | Sand mold (sodium silicate) |
|---|---|---|
| Density (kg/m³) | 7200 | 1600 |
| Thermal conductivity (W/(m·K)) | 35 (at 20°C) | 0.7 |
| Specific heat capacity (J/(kg·K)) | 650 | 1000 |
| Liquidus temperature (°C) | 1510 | — |
| Solidus temperature (°C) | 1455 | — |
| Latent heat (kJ/kg) | 270 | — |
All simulations considered a pouring temperature between 1560°C and 1580°C, consistent with typical steel foundry practice. The pouring process was simulated until the mold was completely filled, and then the solidification was continued until the casting temperature dropped below 1000°C or until complete solidification. Table 4 and Table 5 summarize the boundary and stopping criteria.
| Parameter | Value |
|---|---|
| Solver type | Flow and thermal coupled | −Z coordinate (downward) | Every 5% of filling | Full mold |
| Parameter | Value |
|---|---|
| Stop criterion | Temperature-driven |
| Stop temperature | 1000°C |
| Feeding efficiency | 10% (for default riser) |
| Critical temperature 1 | 1510°C |
| Critical temperature 2 | 1455°C |
| Data storage | Every 5% of solidification |
One of the important pre-processing steps is the placement of tracer particles in the mold. Tracers are massless particles that move with the fluid and allow visual tracking of the flow front. Figure 1 shows the pouring line setup used for a typical automated pouring operation, which is often integrated with the simulation environment to understand the metal delivery into the gating system.
5. Case Study 1: A Heavy Coke Tower Bottom Valve Body
The first practical case is a coke tower bottom valve body weighing nearly nine tons. The component is massive, with a complex geometry involving two large rectangular flanges and a circular flange. The original production route had suffered from recurring shrinkage defects in the flange cores and occasional porosity after machining. I applied the complete simulation workflow to redesign the feeding system.
5.1 Initial Modeling and Simulation
I constructed a three-dimensional solid model of the valve body from the engineering drawing. The model was simplified by ignoring small machining allowances and fillets that do not affect the solidification behavior. The initial casting design included a bottom gating system with two side runners and multiple ingates placed at the lower portion of the casting. No risers were initially attached to study the natural solidification pattern. Figure 2 shows the solid model of the valve body used in the simulation.
After meshing and parameterizing the simulation, I performed a filling analysis with tracer particles. The tracer visualization, as illustrated in Figure 3, revealed that the molten steel entered primarily through the two end ingates, causing the two ends of the valve body to stay hot for a long time. The central portion solidified earlier, while the ends formed isolated hot spots. This is exactly the condition that gives rise to shrinkage casting defects.
The solid fraction plot at 90% solidification is shown in Figure 4. It is evident that the last solidifying regions are located inside the two large flanges. Without any riser, these regions would form large macro-porosity. The Niyama criterion and the hot-spot criterion both indicated a high risk of shrinkage porosity in those end regions. The simulation confirmed that a proper feeding system was absolutely necessary.
5.2 First Modification: Adding Risers and Changing the Gating System
Based on the simulation results, I implemented a revised design. Two open risers were placed directly above the two rectangular flanges, and a blind riser was added above the upper circular flange. In addition, the gating system was changed to a more distributed bottom-gating arrangement. Four ingates were placed along the side walls of the component instead of only at the two ends. This modification reduced the temperature concentration at the ends and promoted a more uniform temperature field.
The filling simulation of the modified design showed a more balanced flow pattern. The filling front advanced from the bottom toward the top without impinging on the core. The solidification sequence changed such that the risers became the last regions to solidify. The risers effectively fed the flange portions, and the Niyama values in the flanges improved considerably. However, the simulation also indicated that the roots of the open risers still contained some dispersed porosity. This was attributed to the insufficient feeding distance near the riser neck.
5.3 Second Modification: Enlarging Risers and Optimizing Chills
To eliminate the remaining micro-porosity at the riser roots, I enlarged the riser dimensions and increased the height of the riser neck. In addition, I replaced the trapezoidal chills originally located under the flanges with longer rectangular chills. The purpose was to sharpen the temperature gradient toward the risers and to establish a more efficient directional solidification path.
The final simulation showed no indication of macro-shrinkage in the cast body. The temperature contours displayed a clean progression of the solidification front from the lower portions toward the risers. The Niyama criterion field was above the critical threshold in all critical sections. The X-ray-like internal view did not reveal any isolated hot spots. After implementing this design in the foundry, the actual castings produced were pressure-tight and passed all ultrasonic inspections. This case demonstrates that casting defects such as shrinkage cavities and porosity can be systematically eliminated by combining simulation with an understanding of solidification physics.
6. Case Study 2: Optimization of a Coke Tower Bottom Valve Cover
The second case involves a valve cover for the same coke tower bottom valve. This component is characterized by a large flange, a smaller flange, and several internal ribs that create serious hot spots where the ribs intersect the flanges. The foundry had experienced low yield due to internal shrinkage in the rib intersections and leakage after machining the large flange. I applied the simulation technique to redesign the process.
6.1 Original Process and Simulated Defect Prediction
The original process used horizontal casting with the large flange oriented vertically. Four risers were positioned on the large flange, one riser on the small flange, and an additional blind riser at a protrusion. No chills were used. The gating system was a bottom type with five ingates directed into the flange roots. A simulation was run using the parameters listed in Table 3 through 5.
The filling and solidification simulation produced the solid/liquid phase distributions shown in Figure 5. At 70% and 90% solidified, it was clear that the rib intersection regions remained liquid after the surrounding areas had frozen. This resulted in isolated liquid pockets that could not be fed by the risers. The Niyama criterion in these regions was far below the critical value, indicating severe micro-shrinkage. Similarly, the lower part of the large flange showed an isolated hot spot in the cross-sectional view, which explained the leakage observed during pressure testing. The small flange itself was sound, but the area adjacent to it exhibited feeding difficulties.
6.2 Process Improvement with External Chills
To eliminate the casting defects, I added external chills in critical locations. Steel chills were placed at the lower parts of the large and small flanges, and smaller chills were applied at the rib intersections. The chills accelerated the solidification in these hot regions, allowing the surrounding sections to solidify almost simultaneously. The riser dimensions were also recalculated to achieve a proper modulus ratio. As shown in Figure 6, the improved design had a much more even solidification field. The rib cross-sections solidified at nearly the same time as the adjacent walls, preventing the formation of isolated liquid pools.
The Niyama criterion after the modification was above the threshold in all monitored regions. Only a slight porosity remained at the far end of the large flange, which was subsequently addressed by adjusting the chill thickness. After implementing the improved process, the castings were machined and subjected to hydrostatic testing. No leakage was observed, and the internal quality met the required specifications.
7. Case Study 3: Solving Leakage Problems in a High-Temperature Plate Valve
The final case is a high-temperature plate valve used in a petroleum refinery. This valve had a long-standing problem of leaking under hydrostatic testing, leading to high scrap rates and even in-service failures. The casting material was CF8M stainless steel (equivalent to 316 stainless), which has a wide freezing range and a high tendency to form micro-porosity. The existing process used four upright flanges, each fed by an open riser, with a central blind riser on top of the valve body. The gating system was a stepped system with ingates at the bottom and at the mid-height parting line.
7.1 Simulated Root Cause Analysis
The simulation of the original process immediately revealed several problems. First, the stepped gating system caused the upper ingates to deliver metal only after the metal level had risen above the runner, leading to an uncontrolled flow. Some metal began to solidify in the upper runner and was later washed into the casting, producing oxide inclusions. Second, the central blind riser was not metallurgically connected to the valve body during solidification. The liquid metal in the riser was isolated from the casting by a solid bridge, so the riser could not feed the thick sections. In fact, the casting was feeding the riser, which is the opposite of what is desired. This reverse feeding created a large shrinkage cavity near the inner ring of the valve.
The cross-sectional Niyama plot, shown in Figure 7, indicated severe porosity in the inner ring and in the center of the four flanges. The temperature field at the end of filling showed a negative temperature gradient, meaning that the riser regions were cooler than the casting sections. This completely undermined the intended feeding direction.
7.2 First Improvement: Simplification and Change of Riser Type
I proposed a series of changes. First, the upper stepped ingates were eliminated, leaving only the bottom gating system. This prevented the premature flow into the upper runner and ensured that all metal entered from the bottom in a controlled manner. Second, the central blind riser was changed to an open riser, which increased its feeding volume and allowed the application of exothermic compounds on the top surface. Third, external chills were added to the lower parts of the flanges to establish a favorable temperature gradient.
The simulation of this revised process showed improved solidification behavior. The four flanges now solidified from their lower parts toward the risers at the top, and the central riser remained liquid until the end. However, a small isolated liquid pocket still existed at the protrusion of the inner ring. This pocket could not be fed by the central riser because the connecting path had already solidified.
7.3 Second Improvement: Local Chills and Riser Neck Design
To address the remaining hot spot at the inner ring, I placed an additional chill along the lower outer edge of the inner ring. This chill, in combination with the central riser, allowed the inner ring to solidify from the bottom upward. The simulation confirmed that the hot spot was completely eliminated. The final solidification pattern indicated that the only remaining liquid at the end was located in the open risers, which is ideal. After implementing this final design in production, the valve bodies passed the hydrostatic pressure test on the first attempt. No leakage was observed, and the internal integrity was confirmed by radiography.
The key lesson from this case is that casting defects often stem from global issues such as unfavorable temperature gradients and improper gating function, not solely from insufficient riser size. Simulation helps to visualize these global patterns and enables targeted corrections.
8. Conclusions and Perspectives
Through the three case studies, I have demonstrated the power of CAE-based casting simulation for the analysis and improvement of valve body castings. The main conclusions of this work are as follows:
First, the combination of three-dimensional solid modeling and finite-difference-based casting simulation provides a reliable and efficient method for predicting casting defects such as shrinkage cavities, micro-porosity, and oxide inclusions. The use of tracer particles and solid-fraction plots allows engineers to identify isolated hot spots before production.
Second, the classical principles of riser design, gating systems, and chills remain highly relevant. However, these analytical methods are limited in their ability to handle complex geometries and dynamic filling effects. Simulation extends these principles by giving quantitative, spatially resolved information about the evolution of temperature, flow, and solidification.
Third, each of the three valve body components studied had a different defect mechanism. In the heavy coke tower bottom valve, the defects were caused by the concentration of ingates at the ends. In the valve cover, the rib intersections created isolated hot spots that required chills. In the high-temperature plate valve, the stepped gating system and reverse feeding were the root causes. The capability to diagnose the actual cause and to test corrective measures virtually proved invaluable.
Finally, the direct comparison between simulation predictions and actual production results confirmed the accuracy of the numerical models. All the optimized processes yielded castings free from major casting defects, with a significant reduction in scrap and rework. The implementation of CAE technology in the foundry not only improves product quality but also shortens the development cycle and reduces the cost associated with trial castings. Looking to the future, I believe that further advances in multi-scale modeling and integrated computational materials engineering will make it possible to predict not only macroscopic casting defects but also the resulting microstructure and mechanical properties. This will bring the foundry industry a step closer to fully digital manufacturing.
In summary, the work described in this article confirms that casting CAE is a mature and indispensable tool for modern foundries. It enables the rapid, accurate, and low-cost exploration of the entire casting process. The systematic use of simulation for process validation and optimization is essential for achieving the goal of defect-free, production-ready valve body castings in today’s competitive manufacturing environment.