In the evolving landscape of modern manufacturing, the demand for agile and rapid production systems has never been higher. As a researcher focused on advanced manufacturing technologies, I have been deeply involved in developing solutions that bridge the gap between design and production, particularly in the realm of metal casting. Traditional sand casting, while versatile, often suffers from prolonged lead times and high costs for prototype development, especially for complex components like automotive engine parts. To address this, my work has centered on integrating stereolithography (SLA) rapid prototyping with conventional sand casting processes, culminating in the creation of a comprehensive CAD/CAE system. This system is specifically tailored for rapid sand casting applications, leveraging the precision and speed of SLA to produce resin patterns that replace wooden or metal counterparts. The goal is to streamline the entire casting workflow, from initial design to final validation, thereby accelerating product development cycles for industries such as automotive, where gray iron and aluminum alloys are predominant.
The core innovation lies in the CAD/CAE system’s ability to adapt casting parameters to the unique characteristics of SLA prototypes. Unlike traditional patterns, SLA resin models exhibit distinct behaviors—such as higher surface roughness, brittleness, and thermal limitations—that necessitate recalibration of standard sand casting parameters. Through extensive experimentation, I have refined key factors like draft angles and linear shrinkage rates to ensure optimal performance in sand casting operations. This system not only facilitates efficient design but also incorporates simulation capabilities to predict and mitigate casting defects, making it a robust tool for rapid sand casting. In this article, I will delve into the system’s architecture, module functionalities, underlying technologies, and practical applications, emphasizing how it enhances the efficiency and reliability of sand casting processes.
The CAD/CAE system is built on the Pro/E 4.0 platform, utilizing its native secondary development interface, Pro/Toolkit, alongside VC++ 2005 for custom programming. This foundation allows for seamless integration of design and analysis modules, targeting rapid sand casting for automotive components. The system is divided into two primary subsystems: the CAD subsystem for casting process design and the CAE subsystem for numerical simulation using ProCAST. Communication between these subsystems is achieved through IGES files, ensuring compatibility and data integrity. The CAD subsystem encompasses three critical modules: Casting CAD Model Design, Gating System Design, and Feeding System Design. Each module is engineered based on established casting principles, with modifications to account for SLA prototype specifics. For instance, the draft angles and linear shrinkage rates are adjusted as per Table 1 and Table 2, which summarize the differences between traditional wood patterns and SLA resin patterns in sand casting.
| Pattern Height (mm) | Wood Pattern Draft Angle | SLA Resin Pattern Draft Angle |
|---|---|---|
| ≤ 10 | 4° 00′ | 6° |
| > 10 – 40 | 2° 05′ | 3.33° |
| > 40 – 100 | 0° 55′ | 1.33° |
| > 100 – 160 | 0° 40′ | 1.08° |
| > 160 – 250 | 0° 35′ | 0.92° |
| > 250 – 400 | 0° 35′ | 0.92° |
| > 400 – 630 | 0° 30′ | 0.83° |
These adjustments are crucial because SLA resin patterns have a higher friction coefficient with sand molds and are more prone to breakage during mold release in sand casting. Similarly, the linear shrinkage rates for castings must be modified to account for the inherent shrinkage of the SLA resin during curing. The DSM Somos 14120 resin used in this system has a linear shrinkage rate of 0.08%, which influences the overall casting dimensions. Revised shrinkage rates for common alloys in sand casting are presented in Table 2.
| Casting Type | Free Shrinkage | Restricted Shrinkage |
|---|---|---|
| Aluminum Alloy Castings | 2.0 – 3.0 | 2.3 – 3.3 |
| Small to Medium Gray Iron Castings | 1.8 – 3.0 | 1.9 – 3.3 |
| Large Gray Iron Castings | 1.7 – 2.9 | 1.8 – 3.0 |
The Casting CAD Model Design module allows users to import part geometries, set parting surfaces, and apply these revised parameters. By accessing a built-in database—developed using Access and ADO technology—designers can quickly assign machining allowances, draft angles, and shrinkage rates. The module employs parametric modeling and dimension-driven techniques, where features are modified through Pro/Toolkit functions that manipulate dimension IDs. For example, the draft angle on a vertical wall can be adjusted programmatically to match SLA requirements, ensuring the pattern can be easily withdrawn from the sand mold without damage. This module forms the bedrock for subsequent gating and feeding system designs, all tailored for sand casting.
Moving to the Gating System Design module, it supports both aluminum and gray iron alloys commonly used in sand casting. The module comprises three sub-modules: an analysis calculator, a modeling tool, and a user-defined feature (UDF) assembly interface. The analysis calculator determines key parameters such as casting weight, wall thickness, and gating ratios based on empirical formulas from casting handbooks. For sand casting, the pouring time ($t_p$) is calculated using the formula:
$$ t_p = k \cdot \sqrt{W} $$
where $W$ is the casting weight in kilograms, and $k$ is a coefficient dependent on the alloy and mold material—typically ranging from 1.8 to 2.2 for sand molds. The minimum cross-sectional area ($A_{\text{min}}$) of the gating system is derived from:
$$ A_{\text{min}} = \frac{W}{\rho \cdot v \cdot t_p} $$
Here, $\rho$ is the metal density, and $v$ is the flow velocity, which is controlled to avoid turbulence in sand casting. The gating ratios (e.g., sprue-to-runner-to-gate areas) are set to 1:2:1.5 for pressurized systems in sand casting, ensuring smooth metal flow. The modeling tool then generates 3D gating components—such as sprue, runners, and gates—with parametric shapes (e.g., trapezoidal or circular sections) that can be resized based on calculated dimensions. Finally, the UDF assembly allows these components to be attached to the casting CAD model, creating a fully integrated gating system ready for sand casting simulation.
The Feeding System Design module follows a similar tripartite structure, focusing on riser design to prevent shrinkage defects in sand casting. The analysis calculator computes the riser modulus ($M_r$) using Chvorinov’s rule:
$$ M_r = \frac{V}{A} $$
where $V$ is the volume of the casting section to be fed, and $A$ is its cooling surface area. For sand casting, the riser modulus is often increased by a safety factor (e.g., 1.2) to account for heat loss in sand molds. The module also considers feeding distance limitations, which for gray iron in sand casting can be expressed as:
$$ L_f = 5 \cdot T $$
with $T$ being the casting thickness. Based on these calculations, the modeling tool creates cylindrical or exothermic risers with dimensions stored in a relational database. The UDF assembly then positions risers on hotspots identified through geometric analysis, ensuring efficient solidification in sand casting. This modular approach enables rapid iteration of feeding designs, crucial for optimizing sand casting processes.
The CAE subsystem, powered by ProCAST, validates the CAD designs through numerical simulation. After exporting the CAD model as an IGES file, it is imported into GeoMESH for geometry repair and meshing. The surface and volume meshes are generated, followed by setting boundary conditions—such as heat transfer coefficients for sand molds—and initial conditions like pouring temperature. The simulation solves the governing equations of fluid flow and heat transfer, including the Navier-Stokes equations for momentum conservation:
$$ \rho \left( \frac{\partial \mathbf{u}}{\partial t} + \mathbf{u} \cdot \nabla \mathbf{u} \right) = -\nabla p + \mu \nabla^2 \mathbf{u} + \mathbf{f} $$
and the energy equation for solidification:
$$ \rho c_p \frac{\partial T}{\partial t} = \nabla \cdot (k \nabla T) + L_f \frac{\partial f_s}{\partial t} $$
where $\mathbf{u}$ is velocity, $p$ is pressure, $\mu$ is viscosity, $T$ is temperature, $c_p$ is specific heat, $k$ is thermal conductivity, $L_f$ is latent heat, and $f_s$ is solid fraction. These simulations predict potential defects like shrinkage porosity or cold shuts, allowing designers to refine gating and feeding systems before physical prototyping. The integration of CAD and CAE thus reduces trial-and-error in sand casting, saving time and material costs.
Key technologies underpin this system’s functionality. First, the interface between Pro/E and MFC is implemented using dynamic link libraries (DLLs), enabling the use of MFC dialogs for user interaction within Pro/E. This allows for intuitive input of sand casting parameters through custom menus. Second, database management via ADO objects stores part libraries and process data, facilitating reuse of designs across different sand casting projects. Third, parametric modeling techniques drive feature dimensions programmatically; for instance, a riser’s height and diameter can be updated automatically based on modulus calculations. Fourth, dimension-driven editing uses Pro/Toolkit’s ProDimension class to modify feature IDs directly, enabling real-time adjustments to draft angles or shrinkage allowances specific to sand casting.
To demonstrate the system’s efficacy, I applied it to the sand casting of an automotive engine main bearing cap made of HT200 gray iron. The process began by loading the part geometry into the CAD subsystem, where parting lines were defined, and revised parameters from Tables 1 and 2 were applied. The gating system was designed using the module’s calculator, which output a pouring time of 8 seconds and a sprue base area of 2.5 cm² for the sand mold. The feeding system added two risers with a modulus of 1.8 cm, positioned atop the cap’s thick sections. The complete CAD model was exported as an IGES file and simulated in ProCAST. The results, as shown in the porosity prediction plot, indicated minimal shrinkage defects, validating the design for sand casting. This case underscores how the system accelerates development—reducing design time from weeks to days—while ensuring quality in sand casting.
Further applications include aluminum alloy parts like ZL105 intake manifolds and cylinder heads. For these, the system’s databases provide alloy-specific parameters, such as higher fluidity coefficients for aluminum in sand casting. The gating ratios are adjusted to 1:1.5:1 for non-pressurized systems, minimizing oxidation. In all cases, the integration of SLA prototypes allows for rapid pattern fabrication; the resin patterns are directly used to create sand molds, bypassing traditional pattern-making bottlenecks. This synergy between rapid prototyping and sand casting is pivotal for low-volume production, where cost and speed are critical.
In conclusion, the CAD/CAE system I developed represents a significant advancement in rapid sand casting technology. By tailoring casting parameters to SLA prototype characteristics and embedding simulation-driven validation, it addresses the unique challenges of modern manufacturing. The system’s modular design—encompassing casting model creation, gating, and feeding—ensures comprehensive coverage of the sand casting process. Through practical implementations, I have observed dramatic improvements in design efficiency and first-pass success rates for sand casting. Future work could expand the database to include more alloys or integrate machine learning for automated parameter optimization. Nonetheless, this system already serves as a robust platform for accelerating product development in sand casting, bridging the gap between digital design and physical realization.