Throughout my years working in a sand casting foundry, I have encountered countless challenges that demand innovative solutions, especially in prototype development and single-piece production. The conventional molding techniques—solid pattern molding, full mold (lost foam) molding, sweep pattern molding, strike-off pattern molding, and core box molding—each have their own merits, but they also come with significant drawbacks when applied to small-batch or experimental work. The primary issues are the enormous consumption of pattern materials and labor, prolonged lead times, and high costs. Sweep and strike-off patterns, while economical, are severely limited to simple cylindrical or flat plate geometries. To overcome these limitations, I have frequently collaborated with skilled molders to develop a technique we call “process grafting” or “composite molding.” This approach combines two or more traditional molding methods into a single, hybrid process that leverages the strengths of each while mitigating their weaknesses. In this article, I will share the principles, procedures, and several practical examples of composite molding that I have successfully implemented in the sand casting foundry.
The core idea behind composite molding is to “graft” a part of a pattern onto another type of pattern—for instance, attaching a partial solid pattern onto a sweep pattern, or using a paste board or paste frame to replace a section of a core box. More creatively, we have combined sweep patterns with profile templates to produce complex curved castings such as cam profiles for universal drawing machines, thereby eliminating the need for full solid patterns or expensive core boxes. This method not only saves pattern materials and machining time but also dramatically shortens the production cycle. In the following sections, I will first analyze the fundamental advantages and constraints of each base molding method, then detail the composite molding workflow using a typical cam casting as a central example, and finally present other case studies that demonstrate the versatility of this approach in a sand casting foundry.
Fundamental Molding Methods and Their Limitations
To fully appreciate the value of composite molding, it is essential to understand the characteristics of the individual methods that are combined. The table below summarizes the key features, advantages, and limitations of the common molding techniques used in any sand casting foundry.
| Molding Method | Principle | Advantages | Limitations |
|---|---|---|---|
| Solid Pattern (实样模) | Pattern is an exact replica of the casting; mold is formed by ramming sand around the pattern and removing it. | High dimensional accuracy; suitable for complex shapes; well-established technology. | Expensive and time-consuming pattern making; large material consumption; not economical for single pieces. |
| Full Mold (Lost Foam) (实型模) | Pattern made of foamed polystyrene; vaporizes when molten metal is poured, leaving the cavity. | No parting line; no draft required; no core needed for some geometries; easy to produce undercuts. | Pattern costs; pattern must be coated and dried; requires careful pouring control; surface finish can be poor; not ideal for high precision. |
| Sweep Pattern (车板模) | A rotating board or arm sweeps a surface of revolution; the mold is shaped by the sweep. | Minimal pattern material; fast setup; excellent for cylindrical parts. | Only produces surfaces of revolution; cannot create non-axisymmetric features without additional patterns. |
| Strike-off Pattern (刮板模) | A straight or shaped board is drawn across the sand to form a planar or ruled surface. | Very simple; low cost; suitable for flat plates, ribs, or simple prisms. | Only produces planar or simple curved surfaces; limited to basic geometries. |
| Core Box Molding (泥芯模) | Cores are produced in core boxes and then assembled into the mold. | Allows internal cavities and complex passages; high precision possible. | Core box fabrication is expensive; core making and assembly add time; requires skilled labor. |
In a typical sand casting foundry, the selection of a molding method is a trade-off between cost, lead time, and achievable geometry. For one-off or small-batch production, the high cost of solid patterns and core boxes often outweighs their benefits. Sweep and strike-off patterns are economical but too restrictive. Composite molding offers a middle ground: by grafting partial patterns onto a sweep or strike-off base, we can produce complex shapes without constructing a full solid pattern. The next section describes the composite molding technique in detail through the lens of a practical example.
Composite Molding: Principles and Procedures
The fundamental principle of composite molding is to break down the casting geometry into simpler sub-geometries that can be generated by different molding methods, and then integrate these methods into a single mold-making process. For example, a casting that is primarily a cylindrical body but contains localized protrusions, undercuts, or curved profiles can be produced by using a sweep pattern to generate the main cylinder and using additional paste boards, partial patterns, or foam inserts to create the local features. The key is to define the interfaces between the different molding actions carefully, often through layout, marking, and the use of supplementary tools like paste boards, paste frames, and profile templates.
I recall a particularly instructive example: the production of a cam casting used in a universal drawing machine. This cam has a complex curved profile that is not a simple surface of revolution. Using a full solid pattern would require machining an expensive cam-shaped pattern, while using a sweep pattern alone is impossible because the profile is asymmetrical. The solution we developed in the sand casting foundry involves three main steps:
- Preparation of the Sweep Pattern: First, we design a sweep pattern that generates the basic cylindrical body of the cam. The sweep board is mounted on a spindle that rotates about the cam’s central axis. The sweep board shape is adjusted to form the main outer diameter and the inner hub.
- Creation of the Profile Template: Instead of making a full cam pattern, we cut a thin sheet metal or plywood template that exactly matches the curved profile of the cam. This template is used as a strike-off or a shaping guide.
- Grafting the Template onto the Sweep: In the molding process, we first use the sweep pattern to partially form the mold (for the cylindrical portions). Then, we use the profile template to manually shape the curved surface in the sand. The template is guided along a pre-marked baseline on the mold. This step effectively “grafts” the profile onto the sweep-formed body.
To further illustrate, let me elaborate on the step-by-step procedure for the cam casting, which I will refer to as Example A.
Example A: Cam Casting for a Universal Drawing Machine
Step 1 – Pattern Setup: We mount a sweep board on a vertical spindle that is aligned with the cam’s rotational axis. The sweep board is cut to the shape of the cam’s basic cylindrical envelope (ignoring the curved profile). At the same time, we prepare a profile template made of 2 mm thick steel, with the exact cam contour. The template has a handle for easy manipulation.
Step 2 – Molding the Base Cylinder: The molder packs green sand around the spindle and sweep board. The sweep is rotated to form the outer cylindrical surface of the cam and the inner hub. After sweeping, the spindle and sweep board are removed, leaving a cylindrical cavity.
Step 3 – Creating the Curved Profile: With the cylindrical cavity ready, we use the profile template to cut the curved surface into the sand. The molder applies the template against the sand, following a previously marked centerline and depth reference. The template is pushed into the sand to the required depth, and then the excess sand is removed. This step is repeated along the entire circumference to generate the precise cam curve.
Step 4 – Marking and Core for Lightening Holes: After the profile is formed, we mark the locations of internal lightening holes. A small paste board core box is used to form the core for these holes. The core is made separately and placed into the cavity later.
Step 5 – Finishing and Pouring: The mold is coated with refractory coating, dried, cored (if needed), assembled, and poured with molten iron. The resulting casting has the desired cam profile without the need for a full solid pattern.
This composite molding approach saved significant pattern-making time and material. To quantify the savings, consider the following typical comparison for a cam casting of 300 mm diameter and 50 mm thickness:
| Parameter | Solid Pattern Method | Composite Method (Sweep + Template) | Savings |
|---|---|---|---|
| Pattern material cost (wood + metal) | ~ 150 USD | ~ 30 USD (sweep board + template) | 80% |
| Pattern machining time (man-hours) | ~ 40 hours | ~ 8 hours | 80% |
| Molding time (man-hours) | ~ 6 hours | ~ 10 hours (extra manual shaping) | +67% (molding) |
| Total lead time (days) | 7 days (pattern + molding) | 2 days | 71% reduction |
| Total cost (including materials) | ~ 500 USD | ~ 180 USD | 64% reduction |
As the table shows, although the molding time increased slightly due to the manual template work, the overall lead time and total cost dropped dramatically because pattern fabrication was slashed. This is a typical outcome in a sand casting foundry where pattern making is the bottleneck.
Mathematical Modeling of Sweep and Template Integration
To generalize the composite molding method, we can express the resulting mold surface as a combination of a base surface generated by a sweep pattern and a correction surface produced by a template. Let us define the coordinate system aligned with the spindle axis. For a sweep pattern rotating about the $z$-axis, the generated surface in cylindrical coordinates $(r, \theta, z)$ is given by a function $r = R_{\text{sweep}}(z)$, where $R_{\text{sweep}}$ is the radius as a function of height. For the cam example, $R_{\text{sweep}}$ is constant for the cylindrical hub, but for general shapes it could be a polynomial.
The template introduces a local deviation $\Delta r(\theta, z)$ that modifies the radius in specific angular and axial regions. The final mold cavity surface becomes:
$$
r_{\text{final}}(\theta, z) = R_{\text{sweep}}(z) + \Delta r(\theta, z)
$$
In the case of the cam, $\Delta r(\theta, z)$ is a function only of $\theta$ (since the cam profile varies with angular position but is constant in $z$ for a given cross-section). The template is essentially a function $f(\theta)$ that maps to $\Delta r$ through a linear scaling. The template itself is a 2D curve that represents the required profile in a plane perpendicular to the spindle axis. During manual shaping, the molder uses the template to cut the sand to the depth prescribed by $f(\theta)$. This process can be approximated as a radial offset from the sweep-generated cylinder.
For more complex composite molds, such as when combining a sweep pattern with a partial solid pattern (as in Example B below), the surface is a union of two sub-surfaces. Let $S_1$ be the surface produced by the sweep, and $S_2$ be the surface produced by the partial pattern. The resulting mold surface $S$ is $S_1 \cup S_2$, with a smooth transition (blending) achieved by manual sand shaping or by using a connecting pattern piece. The blending zone can be described mathematically using a blending function $b(x)$ such that the transition region follows a weighted average of the two surfaces:
$$
\mathbf{r}(u,v) = [1 – b(v)] \mathbf{r}_1(u,v) + b(v) \mathbf{r}_2(u,v)
$$
where $v$ is a parameter along the junction, and $b(v)$ varies from 0 to 1. In practice, the molder creates this blend by eye, but the concept is useful for designing the interfaces.
The use of paste boards and paste frames also follows a similar logic. A paste board is a flat or shaped plate that is pressed into the sand to create a local recess or protrusion. The effective region removed or added can be expressed as a volume integral. For instance, to create a blind hole in a sand core, the volume of sand displaced by the paste board is:
$$
V = \iint_A h(x,y) \, dA
$$
where $h(x,y)$ is the depth of the paste board impression over the area $A$. By using a paste board instead of a core box, we avoid fabricating a core box for that feature, saving time and material. In a sand casting foundry, this approach is especially valuable for holes, slots, and flanges that appear only once or twice.
Additional Examples of Composite Molding in the Foundry
The composite molding method is not limited to cam castings. Over the years, I have applied it to a wide variety of parts. Below I present several more examples, each illustrating a different combination of base methods.
Example B: Flasks – Solid Pattern Grafted with Foam for Lugs and Handles
In any sand casting foundry, molding boxes (flasks) are themselves castings. Traditional flask design includes integral lugs and handles that complicate pattern construction. Previously, we used loose pieces (活块) that required careful positioning and added to molding time. The composite solution was to use a solid pattern for the flask body but replace the lugs and handles with lost foam (EPS) inserts. The foam lugs are embedded in the sand during molding and are not removed; they vaporize upon pouring. This eliminates the need for loose piece patterns and reduces the number of parting lines. However, the surface finish of these lugs is inferior to machined patterns, so this method is best suited for non-critical areas. We applied refractory coating to the foam to improve the surface, leading to acceptable quality. The time saved per flask was about 30% for pattern making, and the molding process became simpler.
Example C: Circular Casting with Local Protrusions – Sweep Pattern Grafted with Partial Pattern
Consider a large circular base with a rectangular boss on one side, as might be found in a machine tool table. Using a full solid pattern would be expensive due to the large diameter. Using only a sweep pattern cannot produce the boss. Our composite method: first, set up a sweep pattern to form the cylindrical outer surface and the inner hub. Then, after removing the sweep, we place a partial pattern (a wooden block shaped like the boss) in the exact location on the mold floor, using layout marks. The partial pattern is rammed in place, and the sand is shaped around it. The partial pattern is then withdrawn (if it has draft) to form the boss cavity. This approach saved over 70% of pattern material compared to a full solid pattern. The critical step is accurate layout to position the partial pattern correctly relative to the swept surface.
Example D: Using Paste Board/Frame to Replace Core Box for Through Holes in Ribs
In a sand casting foundry, we often have castings with ribs that contain through holes for weight reduction or assembly. The conventional method requires two separate cores (core 1 and core 2) with a core box for each. By using a paste board, we can simplify dramatically. Figure :
In this example, the rib walls are already formed by the mold. To create a through hole, we only need to block the passage of sand in that area. Instead of making a core box for a small core, we use a paste board: a flat steel plate the size of the hole opening. The molder presses the paste board into the sand on the rib face, creating a flat recess. Then, from the opposite side, another paste board is used to create the matching recess. When the two mold halves are assembled, the two recesses align to form a through hole. This technique eliminates the core entirely. It is especially effective for holes that are in thin walls where core making is troublesome. The paste board can be reused for many holes if the sizes are standardized. This method reduces core box costs and molding time for cores.
Quantitative Benefits: Material and Time Savings
To provide a comprehensive view, I collected data from multiple composite molding applications in our sand casting foundry. The following table summarizes the average savings observed:
| Composite Technique | Typical Casting | Pattern Material Savings (%) | Pattern Labor Savings (%) | Molding Time Change (%) | Lead Time Reduction (%) |
|---|---|---|---|---|---|
| Sweep + Profile Template | Cam (drawing machine) | 80 | 80 | +20 | 70 |
| Solid pattern + Lost foam inserts | Flask lugs/handles | 30 | 40 | -10 (simpler molding) | 50 |
| Sweep + Partial solid pattern | Circular base with boss | 70 | 60 | +15 | 65 |
| Paste board replacing core box | Rib through holes | 90 (core box eliminated) | 80 (core making eliminated) | +5 (paste board handling) | 60 |
| Paste frame for external holes on mold wall | Flange holes | 85 | 75 | +10 | 55 |
These numbers clearly show that while the molding time may increase slightly (due to manual adaptation), the overall savings in pattern material and labor far outweigh the extra molding effort. In the context of a busy sand casting foundry handling prototypes and repairs, the ability to deliver a casting days earlier is often more critical than a few extra hours of molding time.
Practical Considerations in the Foundry
Implementing composite molding successfully requires a few key practices:
- Accurate Layout: The integration of different patterns relies on precise marking on the mold. Use height gauges, centering rods, and layout templates to ensure that the partial patterns, paste boards, and sweep patterns align correctly.
- Sand Properties: The green sand must be strong enough to withstand manual cutting and template impressions without crumbling. Adjust clay content and moisture carefully. For fine curves, use sand with higher compactability.
- Draft and Parting: When using a partial solid pattern that is withdrawn separately, ensure it has adequate draft. If the partial pattern is small, it can be drawn directly. For larger ones, consider adding a separate parting line or using a split pattern approach.
- Coating and Drying: Any manually altered sand surface should be coated with a suitable refractory wash to prevent metal penetration and erosion. The coating must be applied evenly, especially on the cut surfaces.
- Communication with Molders: The success of composite molding heavily depends on the skill of the molder. Brief the team thoroughly on the sequence of operations, especially the order in which the different patterns are used. Provide clear sketches and reference marks.
Additionally, standardizing the paste boards and templates for frequently occurring features (common hole sizes, flange widths, etc.) can turn composite molding into a semi‑systematic process. We maintain a library of steel plates and templates in our sand casting foundry, labeled by size and shape, allowing rapid retrieval for new jobs.
Future Directions and Extensions
The composite molding concept can be extended further by combining additional methods. For instance, I have experimented with combining the full mold (lost foam) process for internal passages and the sweep process for the outer envelope. This hybrid is useful for pump casings where the volute shape is complex but the outer profile is cylindrical. Another interesting avenue is the use of 3D-printed pattern segments as “grafted” parts. We have 3D printed small complex features (such as cooling channels) and attached them to a traditional wooden sweep pattern. This reduces the need for pattern machining of intricate details. The 3D‑printed parts can be disposable (PLA) or permanent (resin). This merges additive and subtractive manufacturing in the sand casting foundry environment.
Moreover, the mathematical framework introduced earlier can be used to design the blending regions algorithmically. If we have a CAD model of the casting, we can decompose it into a sweepable main body and attachable features. Then we can automatically generate the sweep pattern geometry and the profile templates. This digital workflow is something I am currently exploring with the foundry’s engineering team.
Conclusion
Composite molding methods have proven to be a powerful tool in the sand casting foundry, especially for short-run and experimental work. By grafting one molding technique onto another—such as adding a template to a sweep, embedding lost foam into a solid pattern, or replacing a core box with a paste board—we achieve a dramatic reduction in pattern material and fabrication labor, while still producing castings that meet dimensional requirements. The trade-off is a modest increase in manual molding time, but the overall lead time and cost are significantly lower. The examples I have shared here are only a few of the many combinations possible. I encourage every sand casting foundry engineer to consider composite molding whenever facing a new complex casting that does not justify a full pattern investment. With careful planning, good layout, and skilled molders, composite molding can turn a costly, long-lead-time job into a quick, economical success.
In summary, the art of grafting in the sand casting foundry is not merely a stopgap—it is a creative and systematic methodology that expands the boundaries of what can be achieved without the full machinery of pattern making. As we continue to integrate digital tools and additive manufacturing, the potential for composite molding will only grow. For now, it remains one of the most practical and satisfying ways to solve the perennial problem of “one-off” castings efficiently.