In the sand casting foundry, the use of filters has become a mature technique for purifying molten metal. Over years of practice, I have observed that filters effectively reduce inclusions and improve casting quality. However, to achieve optimal purification, many aspects require further investigation. In this paper, I discuss the origins of inclusions, the purification mechanisms of filters, the types and selection of filters, their placement methods, and the hidden hazards of filter residues. Through systematic experiments in our sand casting foundry, I demonstrate how long-term filter usage degrades reused sand and return materials, and I propose countermeasures. The findings aim to help sand casting foundry engineers apply filter technology more rationally while maintaining the foundry ecosystem.
Origins of Inclusions in Molten Metal
In my long-term production experience in the sand casting foundry, inclusions originate from several main sources:
- Raw material contamination: During melting, impurities from various charge materials, along with physical-chemical reactions, generate numerous metallic and non-metallic inclusions. The use of unshot-blasted return material, severely rusted pig iron, and scrap steel greatly exacerbates this issue.
- Spheroidizing/vermicularizing reactions: Ductile iron and compacted graphite iron produce more inclusions than gray iron. The spheroidizing reaction generates magnesium-rich slag, which consists of sulfide particles combined with magnesium silicate microfilms—high-melting-point inclusions (≥2000 °C). The reaction between MgO and SiO2 under strong Mg deflagration forms magnesium silicate.
- Nodulizer composition: When the nodulizer (or vermicularizer) has high Si content, the formation of magnesium silicate is proportional to the Si increase.
- MnS inclusions: MnS forms in large quantities before and after spheroidizing treatment, depending on sulfur content changes.
- Refractory erosion: Refractory materials from furnace linings and ladle linings detach into the melt, and metal oxides adhering to refractories remelt to form oxide slags.
Purification Mechanisms of Filters
Placing a filter at the bottom of the pouring cup, in the runner, or in the sprue allows the molten metal to momentarily stop, enabling low-density slag inclusions to float and separate. The irregular or regular mesh holes also create deep adsorption effects. Two main mechanisms operate:
| Mechanism | Description |
|---|---|
| Surface filtration (two‑dimensional) | Common to all filters. Large slag inclusions are captured on the filter surface. Captured inclusions form finer filter eyes, enhancing subsequent capture. However, as pouring temperature drops, viscosity rises, and mesh clogging may occur, worsening filling conditions. |
| Internal filtration (three‑dimensional) | Unique to ceramic foam filters. The three‑dimensional channel structure adsorbs and blocks inclusions. The solid ceramic skeleton has a “coherent correspondence” with inclusions, causing adsorption. Inclusions that pass the surface layer repeatedly contact the large‑surface‑area skeleton and are adhered. Additionally, the SiC on the skeleton surface oxidizes to glassy SiO2, which softens when heated, further trapping inclusions. |
The filtration efficiency can be expressed by the ratio of filtered area to runner cross‑section. A common guideline is:
$$ A_{\text{filter}} = k \cdot A_{\text{runners}} $$
Where \(A_{\text{filter}}\) is the filter area, \(A_{\text{runners}}\) is the total cross‑sectional area of the runner, and \(k\) is typically between 2 and 3 to maintain adequate flow rate.
Classification and Characteristics of Filters
Three filter types are widely used in the sand casting foundry: fiber filters, ceramic foam filters, and honeycomb ceramic straight‑hole filters. Their properties differ significantly.
| Property | Value |
|---|---|
| Working temperature (°C) | 1400–1550 |
| Softening point (°C) | >1750 |
| Continuous working time (min) | 5–10 |
| Room temperature tensile strength (4 strands, MPa) | >80 |
| Gas evolution (cm³/g) | 35–44 |
| Property | Value |
|---|---|
| Open porosity (%) | 80–90 |
| Refractoriness (°C) | >1500 |
| Flexural strength (MPa) | >0.8 |
| Compressive strength (MPa) | >0.9 |
| Thermal shock resistance (1100°C, cycles) | >6 |
Fiber filters are lightweight, cheap, and easy to cut, but their slag removal effect is moderate due to lack of deep adsorption. Ceramic foam filters excel in deep filtration but are more expensive. Honeycomb straight‑hole filters offer high strength and stable flow, resisting clogging even with dirty melts.
Selection and Placement of Filters
In my sand casting foundry, selecting the right filter involves considering the alloy type, filter capacity (pouring weight), expected filtration effect, cleanliness of charge materials, inclusion content and nature, melt fluidity, and filling ability. Using a filter typically reduces pouring speed by 10%–25%. For thin‑walled complex castings (e.g., shells, exhaust manifolds), large‑mesh, high‑porosity filters are preferred.
Placement guidelines:
- Use semi‑open or open gating systems to avoid turbulence after the filter.
- For multiple cavities: place one filter below the pouring cup; when multiple filters are needed horizontally at the sprue base, group them.
- For single cavities: place the filter vertically on the runner near the casting.
- Ensure the filter area is 2–3 times the cross‑sectional area of the runner or sprue.
- Design the filter seat about 1 mm larger than the filter to allow clearance; include a sand trap groove to prevent sand inclusion.
- In vertical parting processes (e.g., DISA line), use a pre‑positioned filter frame integrated with the core setter.
The flow reduction factor can be estimated by:
$$ \Delta Q = \left( 1 – \frac{A_{\text{open}}}{A_{\text{runner}}} \right) \times 100\% $$
where \(A_{\text{open}}\) is the total open area of the filter mesh. Typically, \(A_{\text{open}}\) should be at least twice the runner area to keep velocity drop below 20%.
Hazards and Mitigation of Filter Residues
Long‑term, large‑scale use of filters in a sand casting foundry leads to undesirable consequences. After pouring, filters are severely burned and fragmented; their residues enter the recycled sand and return material, degrading the sand system and increasing melting difficulty. In my experiments, I designed a comparative study between two DISA lines in our sand casting foundry.
Experimental Design
Line 1: Long‑term heavy use of filters in complex castings.
Line 2: Simple castings without filters.
Sand test: Both lines used their own reclaimed sand (same moisture), same bentonite and coal dust, same mixing procedure. During production of the same casting (left and right cavities), five sand samples were taken every 30 minutes (at 30, 60, 90, 120, 150 min) from the belt before the molding machine. Properties were measured.
Melting test: Both lines used their own shot‑blasted return material, same pig iron, scrap, same melting procedure. For five heats per line, samples were taken for chemical analysis, and the amount of deslagging agent used per heat was recorded.
Results
| Sample | Wet Compressive Strength (MPa) | Wet Tensile Strength (MPa) | Crack Strength (MPa) | Compactability (%) | Permeability | Friability Index (%) |
|---|---|---|---|---|---|---|
| 1 | 0.200 | 0.021 | 0.031 | 38 | 120 | 80 |
| 2 | 0.187 | 0.021 | 0.031 | 36 | 120 | 78 |
| 3 | 0.208 | 0.0235 | 0.030 | 38 | 135 | 78 |
| 4 | 0.191 | 0.0205 | 0.031 | 37 | 130 | 80 |
| 5 | 0.189 | 0.0225 | 0.032 | 38 | 120 | 76 |
| Sample | Wet Compressive Strength (MPa) | Wet Tensile Strength (MPa) | Crack Strength (MPa) | Compactability (%) | Permeability | Friability Index (%) |
|---|---|---|---|---|---|---|
| 1 | 0.170 | 0.017 | 0.028 | 33 | 130 | 60 |
| 2 | 0.170 | 0.018 | 0.028 | 34 | 140 | 60 |
| 3 | 0.162 | 0.017 | 0.028 | 34 | 145 | 56 |
| 4 | 0.158 | 0.016 | 0.027 | 32 | 130 | 60 |
| 5 | 0.153 | 0.016 | 0.027 | 30 | 100 | 54 |
The data show significant deterioration: wet compressive strength dropped by about 15–20%, compactability decreased, and friability index fell sharply. The filter residues in the sand act as contaminants, reducing bonding effectiveness.
| Heat | Line 2 (no filter) (kg) | Line 1 (with filter) (kg) |
|---|---|---|
| 1 | 8.00 | 11.00 |
| 2 | 8.00 | 10.50 |
| 3 | 8.00 | 11.40 |
| 4 | 8.00 | 10.10 |
| 5 | 8.00 | 10.50 |
| Total | 40.00 | 53.50 |
Line 1 consumed 33.75% more deslagging agent on average (10.70 kg vs 8.00 kg per heat), indicating higher slag load from filter residues.
| Heat | C | Si | Mn | P | S | Cu |
|---|---|---|---|---|---|---|
| 1 | 3.840 | 1.900 | 0.210 | 0.034 | 0.024 | 0.140 |
| 2 | 3.820 | 1.970 | 0.200 | 0.037 | 0.026 | 0.150 |
| 3 | 3.800 | 1.930 | 0.190 | 0.039 | 0.023 | 0.120 |
| 4 | 3.810 | 1.950 | 0.210 | 0.038 | 0.025 | 0.140 |
| 5 | 3.790 | 1.890 | 0.220 | 0.036 | 0.024 | 0.130 |
| Heat | C | Si | Mn | P | S | Cu |
|---|---|---|---|---|---|---|
| 1 | 3.820 | 2.100 | 0.220 | 0.036 | 0.031 | 0.130 |
| 2 | 3.830 | 2.110 | 0.230 | 0.033 | 0.037 | 0.160 |
| 3 | 3.790 | 2.120 | 0.190 | 0.037 | 0.037 | 0.120 |
| 4 | 3.850 | 2.110 | 0.200 | 0.036 | 0.039 | 0.140 |
| 5 | 3.800 | 2.120 | 0.210 | 0.038 | 0.034 | 0.140 |
Notably, Si and S levels increased in Line 1: Si rose by about 0.17% (average 2.112% vs 1.948%), and S rose by 0.01% (average 0.0356% vs 0.0244%). This suggests that filter residues containing silica and sulfur compounds contaminate the return material, leading to unintentional alloying.
Mitigation Strategies
To reduce the negative impact of filter residues on the sand casting foundry ecosystem, I propose the following approaches:
- Improve filter composition: Develop filters that leave fewer harmful residues, e.g., low‑silica formulations or materials that easily disintegrate into harmless phases.
- Advanced screening: Install hexagonal rotary screens before the sand silo to remove larger filter fragments. Further research into magnetic or density‑based separation of fine residues from sand and return material.
- Minimize filter usage: Whenever casting quality can be guaranteed without filters, avoid using them. Implement condition‑based usage: only apply filters when charge material cleanliness is poor, or for specific high‑demand castings. Use filters in a targeted, periodic manner rather than indiscriminately.
- Sand rejuvenation: Incorporate dedicated sand reclamation steps (e.g., thermal reclamation) to burn off organic residues and break down ceramic fragments.
Conclusions
Filters are indispensable for improving casting soundness in the sand casting foundry, but their long‑term use creates a hidden cost: degradation of the sand system and contamination of return material. My experimental data from two parallel DISA lines quantify the detrimental effects: reduced sand strength, increased deslagging agent consumption, and elevated Si and S in the base iron. To sustain a healthy foundry environment, engineers must balance filtration benefits with residue management. By refining filter materials, improving separation technologies, and adopting selective filter usage, we can achieve both high quality and long‑term sustainability in the sand casting foundry.