The application of refractory coatings onto sand molds and cores is a critical step in the sand casting process. This layer acts as a barrier, preventing direct contact between the molten metal and the sand aggregate, thereby significantly reducing surface defects such as metal penetration, burn-on, and sand inclusion. The quality of this interfacial layer is paramount for achieving the desired surface finish and dimensional accuracy of the final sand casting products. Coatings are categorized based on several factors: the alloy being cast (e.g., steel, iron, aluminum, copper), the carrier liquid (aqueous or solvent-based like alcohols), the casting process, and the application method (brushing, spraying, dipping, flowing).
Among these, brushing remains a prevalent manual application method, particularly for complex cores or low-volume production. A persistent challenge in coating application is controlling its penetration into the porous sand matrix and the subsequent built-up thickness on the surface. Aqueous coatings often suffer from insufficient penetration, leading to poor adhesion, while alcohol-based coatings can exhibit excessive penetration, resulting in an inadequate surface layer. The optimal scenario is a balanced combination of sufficient penetration for strong bonding and adequate surface thickness for effective barrier protection. An unsuitable coating thickness can lead to peeling, cracking, or sintering during metal pouring, directly compromising the yield and quality of sand casting products.
The thickness of the dried coating layer is, therefore, a primary performance indicator. For specific sand casting products and coating types, a target thickness range must be established and consistently achieved in production. This thickness is influenced by several process variables, with the coating’s viscosity (often measured as Baumé degree, °Bé) and the number of application layers being two of the most controllable factors in a brushing operation. A higher Baumé degree indicates a higher solids content and viscosity, which affects flowability, penetration behavior, and final build-up. This study focuses on the brushing application of alcohol-based coatings for steel castings, systematically analyzing the individual and interactive effects of coating Baumé degree and the number of brush coats on the resultant dry coating thickness.
1. Experimental Methodology
The experimental procedure was designed to simulate controlled manual brushing conditions. A standardized alcohol-based refractory coating designated for steel sand casting products was used. The coating was maintained in a continuously stirred container to ensure homogeneity. Prior to each application, a sample was taken to measure its Baumé degree using a calibrated hydrometer, ensuring the viscosity was within the specified target range for that trial.
Test substrates were prepared using standard foundry sand bonded with a conventional resin system, representing typical mold/core surface conditions. The brushing procedure was standardized: a consistent brush type and size was used, and a defined brushing technique was employed to minimize operator-induced variation. After each coat was applied, the alcohol carrier was immediately ignited to achieve rapid drying and fixation of the layer. The substrate was allowed to cool to ambient temperature before any subsequent coat was applied. This cycle of brush-ignite-cool was repeated for the required number of coats.
The independent variables were:
- Number of Brush Coats (N): Varied from 1 to 5.
- Coating Baumé Degree (B): Tested at four levels: 75°Bé, 78°Bé, 80°Bé, and 82°Bé.
The primary dependent variable was the Dry Coating Thickness (T). Due to the volatile nature of the alcohol carrier, measuring wet thickness was impractical; therefore, thickness was measured after the final coat was dried and cooled. Measurements were taken using a precision coating thickness gauge with a non-destructive probe suitable for porous substrates. For each combination of N and B, multiple measurements (n=16) were taken across the test surface, and the average was calculated to account for local irregularities inherent to manual brushing and measurement noise. The target thickness range for typical steel sand casting products was identified as 0.4 mm to 0.7 mm, with larger castings requiring up to 1.0 mm.
2. Results and Theoretical Analysis
2.1 Influence of Brushing Frequency on Coating Thickness
Initial experiments held the Baumé degree constant at 75°Bé while varying the number of coats from 1 to 5. The averaged results are summarized in Table 1.
| Test Sequence | Coating Thickness after 1 Coat (mm) | Coating Thickness after 2 Coats (mm) | Coating Thickness after 3 Coats (mm) | Coating Thickness after 4 Coats (mm) | Coating Thickness after 5 Coats (mm) |
|---|---|---|---|---|---|
| 1 | 0.15 | 0.25 | 0.50 | 0.70 | 0.95 |
| 2 | 0.15 | 0.225 | 0.50 | 0.70 | 0.95 |
| 3 | 0.175 | 0.225 | 0.45 | 0.80 | 0.80 |
| 4 | 0.175 | 0.25 | 0.55 | 0.85 | 0.80 |
| 5 | 0.20 | 0.225 | 0.50 | 0.80 | 0.85 |
| 6 | 0.15 | 0.25 | 0.55 | 0.90 | 1.10 |
| Average Thickness (T̄) | 0.17 | 0.24 | 0.51 | 0.79 | 0.91 |
| Incremental Increase (ΔT) | – | +0.07 | +0.27 | +0.28 | +0.12 |
The data reveals a non-linear relationship. The first coat results in a minimal surface thickness (≈0.17 mm) because a significant portion of the coating penetrates into the interstices of the porous sand matrix. This penetration phenomenon can be modeled conceptually by considering the sand as a capillary network. The depth of penetration (P) in a simplified model is influenced by the coating’s viscosity (η), surface tension (γ), contact angle (θ), and pore radius (r), over time (t), often related by a Lucas-Washburn type equation:
$$ P \propto \sqrt{\frac{\gamma r \cos\theta}{2\eta} t} $$
For the first coat on bare sand, penetration is high, leaving a thin surface film. Upon drying, this first layer partially seals the surface pores. Consequently, during the second application, the permeability of the substrate is reduced. Less coating penetrates, and a greater proportion contributes to surface build-up, though the increase (ΔT=+0.07 mm) is still modest. The third coat application shows the most significant incremental gain (ΔT=+0.27 mm), indicating that the previously deposited layers have created a substantially less permeable surface, causing most of the new coating to adhere on top. Subsequent coats (4th and 5th) continue to add thickness, but the rate of increase begins to stabilize or vary, influenced by the rheology of applying paint onto an already built-up, uneven surface. This trend highlights that for achieving the target thickness of 0.4-0.7 mm for common sand casting products, three coats at this viscosity are sufficient.
2.2 Influence of Coating Baumé Degree on Coating Thickness
Based on the initial finding that three coats often reach the target range, we investigated the effect of varying the Baumé degree (B) for both 3-coat and 5-coat processes. The average thickness values for all tested conditions are plotted and presented in Table 2.
| Baumé Degree, B (°Bé) | Avg. Thickness after 3 Coats, T̄₃ (mm) | Avg. Thickness after 5 Coats, T̄₅ (mm) |
|---|---|---|
| 75 | 0.63 | 0.94 |
| 78 | 0.60 | 1.03 |
| 80 | 0.65 | 1.15 |
| 82 | 0.65 | 1.15 |
The results indicate a distinct interaction effect. For the 3-coat process, the coating thickness remains remarkably stable, averaging around 0.63 mm, regardless of the increase in Baumé degree from 75°Bé to 82°Bé. In contrast, for the 5-coat process, the thickness shows a clear positive correlation with Baumé degree, increasing from 0.94 mm at 75°Bé to 1.15 mm at 80-82°Bé.
This divergence can be explained by the competing mechanisms of penetration and surface deposition, which are functions of both the number of layers (N) and the coating rheology (linked to B). The Baumé degree is directly related to the volumetric solids content (C_v) and the apparent viscosity (η). A simple linear correlation can be assumed over the tested range:
$$ \eta \propto C_v \propto B $$
In the initial coats on a porous surface, the primary material loss mechanism is penetration. A higher viscosity (higher B) can slightly reduce the penetration depth (as per the Washburn equation), potentially leaving more material on the surface. However, this effect seems to be counterbalanced by the reduced flowability and leveling, preventing significant extra build-up in just three cycles.
After multiple coats (e.g., N=5), a continuous, less-permeable film has been established. The dominant physics shifts from penetration-controlled to deposition-controlled. Here, the thickness added per coat (δ) becomes more dependent on the coating’s ability to be retained on the existing surface, which is a function of its viscosity and yield stress. A higher viscosity paint has greater resistance to flow under gravity (slump resistance or anti-sag property), leading to less run-off and a higher retained layer per application. Therefore, the cumulative thickness after many coats is more sensitive to viscosity. We can model the final thickness (T_N) after N coats as a sum:
$$ T_N = \sum_{i=1}^{N} \delta_i(B, S_{i-1}) $$
where \( \delta_i \) is the thickness contributed by the i-th coat, which depends on the Baumé degree (B) and the state/sealing of the substrate from the previous (i-1) coats, \( S_{i-1} \). For i=1, \( S_0 \) is the bare sand with high permeability. For i > 3, \( S_{i-1} \) approximates a sealed surface, and \( \delta_i \) becomes more strongly dependent on B.
2.3 Statistical Distribution and Process Window Analysis
To understand the variability and reliability of the process, the entire dataset of 120 thickness measurements was analyzed for frequency distribution at the key application levels of 3 and 5 coats, irrespective of Baumé degree. The results, crucial for quality control in producing sand casting products, are shown in Table 3 and described below.
| Thickness Range (mm) | Frequency for 3-Coat Process (%) | Frequency for 5-Coat Process (%) |
|---|---|---|
| 0.40 – 0.55 | 11 | 0 |
| 0.55 – 0.65 | 61 | 0 |
| 0.65 – 0.80 | 22 | 15 |
| 0.80 – 0.95 | 6 | 26 |
| 0.95 – 1.10 | 0 | 31 |
| 1.10 – 1.25 | 0 | 16 |
| 1.25 – 1.40 | 0 | 12 |
For the 3-coat process, the thickness is tightly clustered, with 61% of measurements falling within the 0.55-0.65 mm band and 94% within 0.4-0.8 mm. This demonstrates that a 3-coat application is a robust and repeatable method for achieving the standard required thickness for many steel sand casting products (0.4-0.7 mm).
For the 5-coat process, the distribution is wider and shifted towards higher values. The most frequent range is 0.95-1.10 mm (31%), suitable for larger or more demanding castings. However, the spread from 0.7 mm to over 1.4 mm indicates that at a high number of coats, other factors like manual brushing inconsistency and the pronounced effect of viscosity variation play a larger role. This necessitates tighter control over coating viscosity and application technique when a thick coating is specified for specialized sand casting products.
3. Conclusions and Industrial Implications
This investigation into the brushing application of alcohol-based coatings for sand molds and cores yields conclusive and actionable insights for foundry practice aimed at improving sand casting products.
1. Law of Incremental Build-up: Starting from a bare sand surface with a coating of 75°Bé, each additional brush coat increases the dry coating thickness by approximately 0.1 to 0.25 mm, with the most significant jump occurring between the second and third coat as the substrate transitions from porous to sealed.
2. Interaction between Viscosity and Application Number: The influence of coating Baumé degree is not isolated but interacts powerfully with the number of coats applied.
- For a standard 3-coat process, the final dry thickness is largely independent of Baumé degree variations within the 75°Bé to 82°Bé range, consistently delivering about 0.63 mm. This makes the 3-coat process very forgiving and reliable for standard specifications.
- For a 5-coat (or higher) process, the final thickness exhibits a strong positive correlation with Baumé degree, increasing by about 0.3 mm across the tested range. The higher solids content and viscosity enhance the coating’s retention on the already formed film, leading to greater build-up per application cycle.
3. Process Windows for Product Categories:
- Conventional Steel Castings: A 3-coat brushing process with alcohol-based coating, controlled within a common viscosity band (e.g., 75-82°Bé), will reliably produce a coating thickness in the range of 0.55-0.65 mm. This satisfies the requirement for the vast majority of standard sand casting products (0.4-0.7 mm).
- Large or Critical Castings: For products requiring a thicker barrier (e.g., ~1.0 mm), a 4 or 5-coat process is necessary. In this regime, the Baumé degree must be a carefully controlled parameter, as it directly dictates the final thickness. A higher Baumé degree (e.g., 80-82°Bé) is more efficient for achieving greater thickness with fewer coats, but process consistency is vital to manage the increased variability.
Therefore, the selection of brushing parameters is not arbitrary but should be a defined part of the process documentation for specific sand casting products. The findings provide a scientific basis for foundry engineers to standardize coating application: use a robust 3-coat process for general work, and for thick-coat applications, implement controlled viscosity and coat number specifications to ensure the interfacial quality that ultimately protects the metal and defines the surface of the cast product.