Integrative CAD/CAE System for Rapid Sand Casting Based on SLA Prototypes

The modern manufacturing landscape is characterized by a demand for agility and rapid response to market needs. Within this context, a critical challenge for next-generation manufacturing systems is achieving both rapid product design and swift physical realization. The fusion of Rapid Prototyping (RP) technologies with traditional casting methods has given rise to Quick Casting (QC), a powerful approach aimed at drastically shortening new product development cycles and reducing associated costs. Among various RP techniques, Stereolithography (SLA) stands out for its ability to produce high-precision, complex, and smooth-surfaced prototypes from photopolymer resin using a UV laser. Key attributes such as near 100% material utilization, high build speeds, excellent dimensional accuracy, and superior surface finish (reaching up to Ra 3.25 μm) make SLA a prime candidate for integration into casting processes. Consequently, rapid casting based on SLA prototypes has become a preferred route, particularly within industries like aerospace, for the swift procurement of complex metal components.

The combination of SLA technology with conventional sand casting practices forms a potent technique known as SLA-based rapid sand casting. This method is exceptionally well-suited for single-unit or low-volume production runs and is ideal for prototyping and new product development. The fundamental workflow involves designing the mold assembly (pattern), fabricating the pattern and core boxes via SLA, assembling these resin prototypes, creating the sand mold, and finally pouring the metal. The pivotal step in this chain is the rapid and accurate design of the mold assembly. This task must not only accommodate the unique characteristics of SLA-fabricated patterns but also adhere strictly to established foundry principles, knowledge, and experiential standards. Utilizing a dedicated Casting Process CAD system is an effective solution to this challenge, enabling the swift design of parting lines, machining allowances, draft angles, cores, gating systems, risers, and chills.

However, existing commercial casting process CAD systems often fall short in providing tailored support for SLA-based rapid sand casting. Many systems are built upon two-dimensional CAD platforms or are narrowly focused on specific part families (e.g., shafts, discs, valve bodies), frequently lacking comprehensive, adaptable databases. More importantly, SLA resin patterns possess distinct characteristics compared to traditional wood or metal patterns. They are more susceptible to deformation, can be brittle and prone to fracture, and have limited thermal resistance. Therefore, when substituting SLA patterns for conventional wooden or metal ones in sand castings, relevant process parameters must be re-evaluated and adjusted based on the specific properties of the photopolymer and the stereolithography process.

To address this technological gap, this work presents the development of an integrative CAD/CAE system specifically for rapid sand casting processes utilizing SLA prototypes. The system is architected with a focus on the rapid development of automotive engine components, primarily for gray iron and aluminum alloys. Its dual functionality encompasses the agile creation of accurate mold CAD models and the subsequent analysis/verification of these designs through numerical simulation within a CAE subsystem.

The core of the CAD subsystem is structured around three essential modules: the Casting CAD Model Design module, the Gating System Design module, and the Feeding (Riser) System Design module. These modules were developed using Pro/ENGINEER Wildfire 4.0 as the foundational platform, leveraging its native Pro/Toolkit API and Microsoft Visual C++ 2005 for creating a customized, user-friendly interface. The CAE subsystem employs the commercial software ProCAST for simulating mold filling, solidification, and predicting potential defects. Communication between the CAD and CAE worlds is facilitated through the IGES (Initial Graphics Exchange Specification) file format, which acts as a robust and sufficiently detailed neutral data exchange interface.

System Architecture and Module Design

The overarching structure of the developed system is designed to streamline the workflow from initial part geometry to a validated casting process layout. The process begins within the CAD environment, where the part model is enriched with manufacturing attributes. Following the design of gating and feeding systems, the complete assembly is exported for simulation. The results from the CAE analysis inform potential design refinements, creating an iterative, knowledge-driven design loop essential for high-quality sand castings.

Casting CAD Model Design Module

This foundational module is responsible for converting a nominal part design into a manufacturable casting model by applying necessary foundry allowances and parameters. The key parameters include machining allowances, draft angles, linear shrinkage (scale factor), and the specification of minimum castable holes. A central innovation of this system lies in its process-specific database. While standard values for parameters like machining allowances are drawn from established foundry handbooks, critical parameters affecting pattern extraction and final casting dimensions are specifically recalibrated for SLA patterns.

Unlike smooth metal or seasoned wood, the surface of an SLA pattern, despite its relatively good finish, creates higher friction with the sand mold. Coupled with the inherent brittleness of cured photopolymer resin, this necessitates an increase in draft angles to prevent pattern damage during mold assembly and disassembly. Based on empirical testing, the draft angles for SLA patterns were revised upward compared to those recommended for wooden patterns. The comparative data is summarized in the table below.

Pattern Height (mm)	Draft Angle for Wood Pattern	Draft Angle for SLA Resin Pattern
≤ 10	4° 00′	6.00°
>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°

Furthermore, the linear shrinkage allowance for the casting must account for two factors: the contraction of the metal upon cooling and the inherent shrinkage of the SLA resin during its photopolymerization cure. For instance, using a material like DSM Somos 14120 resin with a polymerization shrinkage of approximately 0.08% requires a slight adjustment to the traditional casting shrinkage values. The system’s database incorporates these revised, process-compensated linear shrinkage rates for different alloys and constraint conditions, crucial for achieving dimensionally accurate sand castings.

Casting Material	Free Linear Shrinkage (%)	Hindered Linear Shrinkage (%)
Aluminum Alloy Castings	2.0 – 3.0	2.3 – 3.3
Medium/Small Gray Iron Castings	1.8 – 3.0	1.9 – 3.3
Large Gray Iron Castings	1.7 – 2.9	1.8 – 3.0

The module’s functionality is implemented via Pro/Toolkit functions that programmatically access and modify the dimensional parameters (identified by unique ID tags) of Pro/E solid features, allowing for automated application of these revised allowances to the 3D model.

Gating System Design Module

Separate sub-modules were developed for aluminum and gray iron alloys, reflecting their different fluidity and solidification characteristics. Each gating system module comprises three integrated components:

1. Analytical Calculation Sub-module: This interface allows the designer to input casting parameters such as weight, key wall thickness, and desired number of gates. It then automatically calculates critical gating dimensions based on established empirical formulas from foundry handbooks. Key calculations include the optimal pouring time ($t_p$), the choke (smallest cross-sectional) area ($A_c$), and the areas of other gating components (sprue, runners, gates) based on predefined proportional ratios (e.g., $A_{sprue} : A_{runner} : A_{gate} = 1 : 1.5 : 2$). A typical formula for pouring time for medium-sized castings might be: $$t_p = S \sqrt{W}$$ where $W$ is the casting weight in kg and $S$ is an empirical coefficient based on part thickness and complexity. The choke area is then derived from the fluid flow equation: $$A_c = \frac{W}{\rho \cdot t_p \cdot \mu \sqrt{2gH}}$$ where $\rho$ is metal density, $\mu$ is the discharge coefficient, $g$ is gravity, and $H$ is the effective metal head height.
2. Parametric Modeling Sub-module: Based on the calculated dimensions, this component utilizes Pro/E’s robust parametric and relational modeling capabilities. It selects an appropriate cross-sectional shape (trapezoidal, circular, etc.) for each gating element and generates the 3D solid geometry automatically, ensuring correct volumetric and connective relationships for efficient mold filling in sand castings.
3. User-Defined Feature (UDF) Assembly Sub-module: To promote standardization and reuse, commonly used gating designs can be saved as UDF libraries. This sub-module allows for the quick retrieval and placement of these pre-defined gating components onto the casting model within the assembly environment, significantly speeding up the layout process.

Feeding (Riser) System Design Module

Similar to the gating module, dedicated riser design logic is implemented for aluminum and gray iron. Effective feeding is paramount to producing sound sand castings free of shrinkage porosity. This module also follows a three-tiered architecture:

1. Analytical Calculation Sub-module: The designer inputs data such as casting weight, the volume and surface area of sections to be fed, desired riser count, and feeding efficiency factors. The core calculation is the determination of the required riser modulus ($M_r$). The modulus is defined as the volume-to-cooling-surface-area ratio ($V/A$). A fundamental rule is that the riser modulus must be greater than the modulus of the casting section it is intended to feed: $$M_r > k \cdot M_c$$ where $M_c$ is the casting modulus and $k$ is a safety factor (usually >1). The module calculates the casting modulus and then computes the necessary riser modulus using established methods like the “Caine” or “Wlodawer” relationships, which can be expressed in generalized form: $$M_r = f(M_c, V_c, \Delta T, x…)$$ where $V_c$ is casting volume, $\Delta T$ is solidification temperature range, and $x$ represents other material-specific factors.
2. Parametric Modeling Sub-module: Using the calculated riser modulus, this sub-module accesses a database linking modulus values to standard riser geometries (cylindrical, top-heavy, etc.) and their dimensions (diameter $D$, height $H$). It then automatically generates the 3D solid model of the riser, including necessary features like an exothermic sleeve or insulation lining if specified in the database rules.
3. User-Defined Feature (UDF) Assembly Sub-module: Standard riser designs are stored as UDFs. This allows for the automatic placement and Boolean union of the correctly sized riser onto the hot spot of the casting CAD model, ensuring proper metallurgical contact for effective feeding during the solidification of sand castings.

CAD/CAE Integration Interface

A seamless link between design and analysis is critical. While some CAE packages offer direct translators for specific CAD kernels, the chosen approach here utilizes the standardized IGES format. The CAD subsystem exports the complete mold assembly (casting with allowances, gating, and risers) as an IGES file. This file is then imported into the pre-processing module of the ProCAST suite (MeshCAST). Although this is a universal rather than a direct translation, it offers significant advantages: the repair and surface meshing tools within MeshCAST are highly effective for healing minor geometrical inconsistencies and generating a clean, analysis-ready surface mesh, which is a prerequisite for accurate simulation of sand castings.

Key Implementation Technologies

The development of this integrated system hinged on several advanced software engineering and CAD customization techniques:

• Pro/E and MFC Interface via DLL: To create a familiar Windows-style graphical user interface (GUI), Microsoft Foundation Classes (MFC) were used. Since Pro/Toolkit does not natively support MFC, a Dynamic Link Library (DLL) mode was employed to establish communication bridges between Pro/E, the Pro/Toolkit application, and the MFC dialog boxes, enabling a seamless user experience.
• Database Design and ADO Technology: A comprehensive Microsoft Access database was developed to store all process parameters (revised draft angles, shrinkage rates), standard gating/riser dimension ratios, and empirical coefficients. ActiveX Data Objects (ADO) technology is used within the VC++ code to query this database efficiently, providing the necessary data to drive the parametric models and calculations.
• Parametric and Dimension-Driven Modeling: This is the cornerstone of the CAD automation. All gating and riser components are modeled with constrained dimensions and mathematical relations (e.g., sprue height = pouring basin height + cope height). The system manipulates these models by programmatically changing the driving dimensional parameters via their Pro/Toolkit ID codes, effectively generating custom-sized components on the fly based on analytical results.

Application Workflow and Validation

The practical utility of the system is demonstrated through the development process for a gray iron (HT200) engine main bearing cap, a typical sand casting component. The workflow is as follows:

1. The user launches the system and logs into the CAD subsystem.
2. The nominal bearing cap part file is loaded. The appropriate material (HT200) is selected from the database.
3. Using interactive tools, the parting plane is defined, and the casting orientation is set.
4. The Casting CAD Model Design Module is executed. The system automatically applies the SLA-optimized machining allowances, draft angles (from the revised table), and the compensated linear shrinkage factor. It also identifies and flags features below the minimum castable size.
5. The Gating System Module is invoked. For this iron casting, the user inputs parameters, and the system calculates the pouring time, choke area, and sprue/runner/gate dimensions. A tapered sprue with trapezoidal runners is automatically generated and assembled to the casting.
6. The Feeding System Module is activated. The system calculates the modulus of the casting’s thick sections, determines the required riser size and neck dimensions, and places an exothermic-sleeved cylindrical riser onto the model.
7. The complete assembly, now representing the cavity of the sand mold, is exported as an IGES file.
8. This IGES file is imported into ProCAST’s MeshCAST. After automatic geometry repair and surface mesh generation, a volume mesh is created.
9. In the PreCAST module, boundary conditions (heat transfer coefficients for sand mold), material properties (HT200, sand), pouring temperature, and interfacial resistances are assigned.
10. The solver (ProCAST) runs the coupled thermal-fluid-solidification analysis.
11. Results are reviewed in ViewCAST. The simulation visually displays the filling sequence, temperature gradients, and, crucially, predicts the location of shrinkage porosity using a porosity criterion (e.g., Niyama criterion). The designer can assess if the riser is effectively feeding the critical section or if design modifications are needed.

The iterative use of this CAD/CAE system for components like engine blocks, cylinder heads (ZL105 aluminum), and intake manifolds has proven its effectiveness. It dramatically accelerates the casting process design phase, reduces reliance on physical trial-and-error, and enhances first-time success rates for producing high-integrity sand castings from SLA patterns.

Conclusion

This research successfully details the development and implementation of an integrative CAD/CAE system specifically tailored for rapid sand casting processes that employ Stereolithography (SLA) prototypes. The system addresses a critical niche by providing automated design tools that are cognizant of the unique properties of SLA patterns, necessitating revised standards for key parameters like draft angle and linear shrinkage allowance. By embedding this specialized knowledge within a parametric design framework linked to a powerful simulation engine, the system bridges the gap between digital design and physical manufacture for sand castings. It enables a fast, reliable, and knowledge-driven pathway from concept to a validated casting process layout, significantly boosting efficiency in prototyping and low-volume production scenarios. Future work may focus on expanding the material database, incorporating artificial intelligence for automatic optimization of gating and risering, and enhancing the bidirectional data flow between the CAD and CAE environments for even tighter integration.