I begin with the practical problem that appears whenever an external chill is placed in a sand casting mold. The chill is intended to increase the local cooling rate, refine the microstructure, reduce shrinkage porosity, and improve feeding in heavy-section castings. However, the same chill also creates a difficult interface. Liquid iron can penetrate the gap between the chill and the sand mold, sand can sinter onto the chill surface, and a thin layer of metal can weld to the chill. The result is sand adhesion, metal penetration, difficult cleaning, high rejection risk, and poor chill recovery. In my work, I treat this interface not as a single coating problem but as a coupled problem involving refractory aggregate design, suspension stability, binder chemistry, solvent evaporation, and interfacial adhesion to both sand and metal. The central objective is an alcohol-based coating for external-chill sand casting that can adhere to a furan resin sand mold and to a metallic chill at the same time, while remaining fast-drying, strong, refractory, and easy to strip after solidification.
In sand casting, an external chill is usually a metal block placed in the mold cavity or on the mold surface. During pouring, the chill extracts heat from the liquid metal. The local heat flux can be written as
$$q = hA(T_c – T_m)$$
where \(q\) is the heat flux, \(h\) is the interfacial heat-transfer coefficient, \(A\) is the contact area, \(T_c\) is the casting temperature, and \(T_m\) is the chill temperature. A coating lies inside this heat-transfer path. If the coating is too insulating, the chill loses efficiency. If the coating is too thin or defective, metal penetration and sand burn-on occur. Therefore, the coating must be thin, uniform, refractory, and mechanically stable. In sand casting with external chills, the coating must also survive thermal shock, gas evolution, and the mechanical handling that occurs before pouring.
I selected an alcohol-based route because alcohol-based coatings offer rapid drying, low moisture content, and reduced gas defects compared with many water-based coatings. Yet alcohol-based coatings often have weaknesses: poor suspension stability, weak adhesion to metal, excessive binder gas evolution, and limited wear resistance. My aim is to overcome those weaknesses by combining a modified lithium bentonite suspension agent, a graded bauxite-white corundum refractory aggregate, and a phenolic resin-polyvinyl butyral composite binder. I use orthogonal experimentation and weighted normalization to identify a formulation that is balanced rather than merely optimal in one property. The final coating is then validated in a heavy machine-tool bed casting produced by external-chill sand casting.
1. Process Background and Design Logic
Large thin-wall and heavy-section iron castings often contain geometric hot spots. In sand casting, those hot spots cool more slowly than the surrounding sections. The slow-cooling region remains liquid longer and may not receive enough feed metal. Shrinkage porosity, shrinkage cavities, and local property loss can follow. An external chill is a direct and economical method to increase the local cooling rate. It changes the solidification sequence and encourages directional solidification toward the feeder. However, an external chill is not simply a metal block. It is an engineered interface. The chill surface must contact the mold and the casting in a controlled way. The coating on the chill surface and on the adjacent sand mold determines whether the interface remains clean or becomes a defect source.
When liquid iron contacts a bare chill in sand casting, several events occur almost simultaneously. The iron oxidizes, the chill oxidizes, and the sand binder decomposes. If the coating is discontinuous, iron can enter the sand pores and form a mechanical interlock with the sand. This is sand adhesion or burn-on. If the iron reacts with the chill, a welded layer can form. This is metal penetration or chill welding. Cleaning such defects is expensive. It can also damage the chill, reduce its reuse, and increase the risk of dimensional error. The coating must therefore act as a sacrificial barrier. It must resist wetting by liquid iron, limit chemical reaction, and remain compliant enough to avoid cracking under thermal expansion mismatch.
My design logic is that a single-component coating cannot satisfy all requirements in external-chill sand casting. A highly refractory aggregate may improve resistance to metal penetration but can settle quickly. A high binder content may improve green strength and wear resistance but increases gas evolution and viscosity. A high suspension agent content may improve suspension stability but can raise viscosity and reduce brushing performance. I therefore treat the formulation as a multicomponent system with competing responses. The key responses I measure are condition viscosity, 24 h suspension rate, gas evolution, coating wear mass loss, and brushing grade. I then use range analysis to understand the dominant factors and weighted normalization to select the best overall composition.
2. Raw Materials and Functional Roles
I used bauxite and white corundum as the refractory aggregate. Bauxite has high refractoriness, good thermal stability, and moderate thermal expansion. White corundum has high hardness, good chemical resistance, and strong resistance to acid and basic slags. A mixture of the two provides a refractory skeleton with better packing than either material alone. I graded bauxite into 250 mesh and 400 mesh fractions. The 250 mesh fraction improves permeability and resistance to erosion. The 400 mesh fraction fills the voids between coarser particles, increases packing density, reduces open porosity, and contributes to a smoother casting surface. In sand casting with external chills, this graded structure is important because it reduces the probability that liquid iron will penetrate through the coating.
| Material | Role in the coating | Key characteristics |
|---|---|---|
| Bauxite, 250 mesh | Coarse refractory skeleton | High refractoriness, good permeability, moderate expansion |
| Bauxite, 400 mesh | Fine filler and pore reducer | Fills interstices, increases density, improves surface finish |
| White corundum, 250 mesh | Hard refractory phase | High hardness, acid and alkali resistance, good thermal shock resistance |
| Calcium bentonite | Precursor for modified suspension agent | Montmorillonite-rich, high cation exchange capacity |
| Phenolic resin | Primary binder | High carbon residue, good high-temperature bonding |
| Polyvinyl butyral | Secondary binder and film former | Improves adhesion, flexibility, and green strength |
| Industrial ethanol | Solvent | Fast evaporation, low moisture, rapid ignition drying |
The suspension agent is based on calcium bentonite. Natural calcium bentonite has good swelling in water but poor compatibility with alcohol. If I merely convert it to lithium bentonite, the gel can lose water to ethanol and become unstable. I therefore use a two-step intercalation-grafting modification. The first step is lithium exchange. The second step combines organic intercalation and surface grafting. This creates larger interlayer spacing, weaker interlayer forces, and better compatibility with the alcohol medium. The modified bentonite can form a card-house network that supports the dense refractory particles. The network increases the yield stress and prevents settling without making the coating impossible to brush.
The binder system is a phenolic resin and polyvinyl butyral composite. Phenolic resin provides high-temperature bonding and a carbon residue that helps maintain coating integrity during pouring. Polyvinyl butyral improves adhesion to metal, film flexibility, and green strength. However, both binders decompose at high temperature and release gases. The gas evolution must be controlled because excessive gas can create pinholes and blowholes in the casting surface. In external-chill sand casting, gas can also be trapped at the chill interface, where it is difficult to vent. I therefore treat binder content as a critical variable. The binder must be high enough to bond the coating and adhere to the chill, but low enough to keep gas evolution within a safe range.
| Modification step | Chemical or physical action | Expected effect |
|---|---|---|
| Lithium exchange | Replace interlayer cations with lithium ions | Improves swelling and gel formation |
| Organic intercalation | Insert organic ammonium species and polymer chains | Increases interlayer spacing, reduces aggregation |
| Surface grafting | Attach silane and functional organic groups | Improves compatibility with ethanol and resin |
3. Structural Evidence of Bentonite Modification
I verified the modification by X-ray diffraction, scanning electron microscopy, and expansion index measurements. The basal spacing of the bentonite changed in a way that is consistent with successful intercalation and grafting. The initial \(d_{001}\) spacing was 1.26 nm. After lithium exchange, the spacing decreased slightly to 1.22 nm. This decrease is related to the small ionic radius of lithium and the changes in interlayer hydration. After intercalation with organic species, the spacing increased to 3.12 nm. After grafting, it stabilized at 3.09 nm. The increase in basal spacing indicates that organic molecules entered the interlayer region and expanded the galleries. The slight decrease after grafting suggests that the grafted chains partly reorganize the interlayer structure and cross-link the surfaces.
The Bragg relation used for the calculation is
$$2d\sin\theta = n\lambda$$
where \(d\) is the interlayer spacing, \(\theta\) is the diffraction angle, \(n\) is the order of diffraction, and \(\lambda\) is the X-ray wavelength. The measured shift toward lower angles corresponds to a larger \(d\) spacing. This is a direct structural signature of intercalation. The scanning electron microscopy observations are consistent with this interpretation. The natural bentonite appears dense, smooth, and strongly aggregated. The modified bentonite retains the lamellar morphology but becomes looser, rougher, and more porous. The increased surface area and pore volume help the modified bentonite disperse in ethanol and interact with the binder. The expansion index also increases from 5.8 mL/g to 20.1 mL/g. This is a large change and confirms that the modification improves swelling and interlayer accessibility.
| Property | Unmodified bentonite | Modified bentonite | Interpretation |
|---|---|---|---|
| Basal spacing \(d_{001}\) | 1.26 nm | 3.09-3.12 nm | Organic species expanded the interlayer galleries |
| Expansion index | 5.8 mL/g | 20.1 mL/g | Improved swelling and dispersion |
| Morphology | Dense, smooth aggregates | Looser, rougher lamellae | Higher accessible surface and porosity |
| Alcohol compatibility | Poor | Improved | Better suspension stability in ethanol |
The suspension stability of a coating depends on the balance between gravitational settling and colloidal network strength. For a spherical particle in a viscous liquid, the Stokes settling velocity is
$$v_s = \frac{2r^2(\rho_p – \rho_f)g}{9\eta}$$
where \(r\) is the particle radius, \(\rho_p\) is the particle density, \(\rho_f\) is the fluid density, \(g\) is gravitational acceleration, and \(\eta\) is the fluid viscosity. A high suspension rate requires either small particles, a small density difference, a high viscosity, or a yield stress network. In an alcohol-based coating for sand casting, increasing viscosity without limit is not acceptable because the coating must be brushable and must penetrate the sand surface. The modified bentonite solves this problem by forming a thixotropic network. The network has a yield stress at rest but breaks down under shear during brushing. This is why the coating can remain suspended for 24 h while still being easy to apply.
I describe the flow behavior with a general shear-thinning model:
$$\tau = \tau_0 + K\dot{\gamma}^n$$
where \(\tau\) is shear stress, \(\tau_0\) is yield stress, \(K\) is consistency index, \(\dot{\gamma}\) is shear rate, and \(n\) is the flow index. For a shear-thinning coating, \(n < 1\). The yield stress supports the refractory particles at rest, while the shear-thinning index allows the coating to flow during brushing. The modified bentonite provides both the yield stress and the shear-thinning character. This is important in external-chill sand casting because the coating must not drip from vertical chill surfaces or from the sand mold wall.
4. Coating Preparation and Testing Methods
I prepared the coating in several stages. First, I dry-blended the bauxite and white corundum fractions in a wheel mixer. The mass ratio of bauxite to white corundum was fixed at 4:1. The total refractory aggregate mass was 100 parts. I then added the pre-prepared modified bentonite slurry, the phenolic resin solution, the polyvinyl butyral solution, and additional ethanol. I mixed the paste for 2.5 to 3 h until the components were uniformly distributed. After mixing, I discharged the paste and diluted it with ethanol to a density of \(1.60 \pm 0.02\) g/cm\(^3\). This density gives a good balance between solid loading and brushability. If the density is too low, the coating is thin and may not cover the surface. If the density is too high, the coating is viscous and may not wet the sand or the chill evenly.
| Preparation stage | Operation | Purpose |
|---|---|---|
| Dry blending | Mix graded bauxite and white corundum | Create a uniform refractory skeleton |
| Slurry addition | Add modified bentonite slurry | Build suspension network |
| Binder addition | Add phenolic resin and PVB solutions | Provide green and fired strength |
| Wet mixing | Mull for 2.5-3 h | Break agglomerates and wet particles |
| Dilution | Adjust with ethanol to target density | Control viscosity and coating thickness |
For coating tests, I used furan resin sand specimens and metal block specimens. The sand specimens were cylindrical, prepared by impact compaction. The metal specimens represented the external chill surface. I applied two brush coats to each specimen and allowed the coating to stand for 2 min before observing vertical flow. I graded brushing performance as follows: Grade I means both the sand and metal surfaces have a uniform coating, no brush marks, and a thickness between 0.5 and 1.2 mm. Grade II means both surfaces are basically uniform with slight brush marks and a thickness between 0.5 and 1.2 mm. Grade III means at least one surface has uneven thickness, local thickness below 0.5 mm or above 1.3 mm, or obvious brush marks. This grading is important for external-chill sand casting because uneven coating thickness causes uneven chilling and can create local defects.
I measured condition viscosity with a flow cup, 24 h suspension rate by settling observation, gas evolution by a standard gas evolution apparatus, and wear resistance by a coating abrasion test. The gas evolution per unit mass is
$$V_g = \frac{V_t – V_0}{m_c}$$
where \(V_g\) is the gas evolution per gram, \(V_t\) is the total gas volume, \(V_0\) is the blank gas volume, and \(m_c\) is the coating mass. A low gas evolution is desirable because it reduces the risk of gas porosity in sand casting. However, a very low binder content can weaken the coating and reduce adhesion to the external chill. The optimization must therefore balance gas evolution with strength and adhesion.
5. Orthogonal Experimental Design
I selected three factors for the orthogonal experiment: the mass ratio of 250 mesh bauxite to 400 mesh bauxite, the modified suspension agent content, and the composite binder content. I coded the factors as A, B, and C. Factor A had three levels: 60:20, 53:27, and 40:40. Factor B had three levels: 3%, 4%, and 5% by mass of refractory aggregate. Factor C had three levels: 2%, 4%, and 6% by mass of refractory aggregate. The binder itself was a phenolic resin and polyvinyl butyral mixture with a mass ratio of 2.5:1. The alcohol solvent content was adjusted during preparation to reach the target density. I used an \(L_9(3^3)\) orthogonal array. This design allows me to evaluate three factors at three levels with only nine experiments, while still estimating the main effects.
| Factor | Meaning | Level 1 | Level 2 | Level 3 |
|---|---|---|---|---|
| A | Mass ratio of 250 mesh bauxite to 400 mesh bauxite | 60:20 | 53:27 | 40:40 |
| B | Modified suspension agent content | 3% | 4% | 5% |
| C | Composite binder content | 2% | 4% | 6% |
The response variables were condition viscosity, 24 h suspension rate, gas evolution, coating wear mass loss, and brushing grade. For brushing grade, I converted the qualitative grades into numerical values: Grade III = 3, Grade II = 2, and Grade I = 1. Lower values are better for brushing grade, viscosity, gas evolution, and wear mass loss. Higher values are better for 24 h suspension rate. This conversion allows range analysis and weighted normalization.
6. Orthogonal Experimental Results
The nine experimental formulations produced a range of properties. The condition viscosity varied from 5.0 s to 8.3 s. The 24 h suspension rate varied from 88% to 95%. The gas evolution varied from 11.0 mL/g to 15.9 mL/g. The wear mass loss varied from 0.049 g to 0.070 g. The brushing grade varied from Grade I to Grade III. These results confirm that the three factors interact strongly with the coating performance. No single formulation is best in every property. For example, a high binder content improves wear resistance but increases gas evolution and viscosity. A high suspension agent content improves suspension stability but can increase viscosity. A finer aggregate fraction improves surface finish and packing density but can reduce permeability if used excessively.
| Run | A | B | C | Condition viscosity (s) | 24 h suspension rate (%) | Gas evolution (mL/g) | Wear mass loss (g) | Brushing grade |
|---|---|---|---|---|---|---|---|---|
| 1 | A1 | B1 | C1 | 5.0 | 88 | 11.0 | 0.067 | III |
| 2 | A1 | B2 | C2 | 6.5 | 94 | 13.4 | 0.056 | I |
| 3 | A1 | B3 | C3 | 7.3 | 91 | 15.9 | 0.052 | II |
| 4 | A2 | B1 | C2 | 6.2 | 90 | 13.8 | 0.063 | I |
| 5 | A2 | B2 | C3 | 8.3 | 93 | 14.2 | 0.049 | III |
| 6 | A2 | B3 | C1 | 5.9 | 95 | 12.6 | 0.069 | I |
| 7 | A3 | B1 | C3 | 8.0 | 89 | 13.5 | 0.050 | III |
| 8 | A3 | B2 | C1 | 6.0 | 94 | 12.0 | 0.070 | II |
| 9 | A3 | B3 | C2 | 7.1 | 93 | 13.2 | 0.058 | II |
7. Range Analysis and Factor Significance
I used range analysis to estimate the main effect of each factor. For each factor and each level, I calculated the average response \(k\). The range \(R\) is the difference between the maximum and minimum \(k\) values for that factor. A larger \(R\) means that the factor has a stronger influence on the response. The results show that the dominant factor depends on the response. For condition viscosity, the binder content is the most important factor, followed by the aggregate ratio and then the suspension agent content. For 24 h suspension rate, the suspension agent content is the most important factor, followed by the aggregate ratio and then the binder content. For gas evolution, the binder content is the most important factor, followed by the suspension agent content and then the aggregate ratio. For wear resistance, the binder content is again the most important factor, followed by the aggregate ratio and then the suspension agent content. For brushing grade, the binder content is the most important factor, while the aggregate ratio and suspension agent content have similar but smaller effects.
| Response | Factor A \(R\) | Factor B \(R\) | Factor C \(R\) | Order of influence |
|---|---|---|---|---|
| Condition viscosity | 0.76 | 0.53 | 2.24 | C > A > B |
| 24 h suspension rate | 1.7 | 4.7 | 1.3 | B > A > C |
| Gas evolution | 0.63 | 1.13 | 2.66 | C > B > A |
| Wear mass loss | 0.0020 | 0.0017 | 0.0184 | C > A > B |
| Brushing grade | 0.66 | 0.66 | 1.34 | C > A ≈ B |
The range analysis reveals a central trade-off in alcohol-based coatings for external-chill sand casting. The composite binder increases viscosity, wear resistance, and adhesion. At the same time, it increases gas evolution. The suspension agent increases suspension stability but also contributes to viscosity. The aggregate ratio controls packing density and surface quality. A coarse aggregate improves permeability but can leave larger pores. A fine aggregate improves packing and surface finish but can reduce permeability and increase binder demand. The optimum formulation must therefore be a compromise.
For condition viscosity, the best level is C1, which is the lowest binder content. For 24 h suspension rate, the best level is B3, which is the highest suspension agent content. For gas evolution, the best level is C1. For wear resistance, the best level is C3, which is the highest binder content. For brushing grade, the best level is C2, with A2 and B3 also performing well. If I selected each response independently, I would obtain contradictory formulations. This is why I use weighted normalization.
8. Weighted Normalization and Optimal Formulation
I assigned weights to the responses according to their importance in external-chill sand casting. The weights are 10% for condition viscosity, 25% for 24 h suspension rate, 20% for gas evolution, 10% for wear resistance, and 35% for brushing grade. The high weight on brushing grade reflects the practical requirement that the coating must adhere uniformly to both the sand mold and the external chill. The high weight on suspension rate reflects the need for a stable coating during storage and application. The gas evolution weight is also significant because gas defects are difficult to remove in sand casting. The viscosity and wear weights are lower but still necessary for handling and durability.
For positive indicators, I used the normalization
$$x’_{ij} = \frac{x_{ij} – \min_j x_{ij}}{\max_j x_{ij} – \min_j x_{ij}}$$
For negative indicators, I used the normalization
$$x’_{ij} = \frac{\max_j x_{ij} – x_{ij}}{\max_j x_{ij} – \min_j x_{ij}}$$
Here, \(x_{ij}\) is the original value of response \(j\) in run \(i\), and \(x’_{ij}\) is the normalized value between 0 and 1. A higher normalized value is always better. The composite score is
$$S_i = \sum_{j=1}^{m} w_j x’_{ij}$$
where \(w_j\) is the weight of response \(j\), and \(m\) is the number of responses. I calculated the composite score for each of the nine runs. The highest score corresponds to the best overall balance.
| Run | Formulation | Weighted normalized score |
|---|---|---|
| 1 | A1B1C1 | 18.24 |
| 2 | A1B2C2 | 19.81 |
| 3 | A1B3C3 | 18.13 |
| 4 | A2B1C2 | 18.76 |
| 5 | A2B2C3 | 18.52 |
| 6 | A2B3C1 | 20.28 |
| 7 | A3B1C3 | 17.70 |
| 8 | A3B2C1 | 19.79 |
| 9 | A3B3C2 | 19.44 |
The optimal formulation is A2B3C1. This corresponds to a 250 mesh to 400 mesh bauxite mass ratio of 53:27, a modified suspension agent content of 5%, and a composite binder content of 2%. The phenolic resin to polyvinyl butyral mass ratio remains 2.5:1. This formulation achieves a low viscosity, a high suspension rate, and excellent brushing performance. Its gas evolution and wear resistance are not the absolute best among all runs, but they remain within the safe range for sand casting. The measured properties of the optimal coating are a 24 h suspension rate of 95%, a gas evolution of 12.6 mL/g, a brushing grade of Grade I, and excellent wear resistance.
| Property | Optimal coating value | Acceptance requirement | Assessment |
|---|---|---|---|
| Condition viscosity | 5.9 s | 5.5-12 s | Pass |
| 24 h suspension rate | 95% | ≥93% | Pass |
| Gas evolution | 12.6 mL/g | <20 mL/g | Pass |
| Wear mass loss | 0.069 g | <0.5 g | Pass |
| Brushing grade | Grade I | Grade I or II | Pass |
9. Mechanism of Performance Improvement
The improved performance of the coating comes from several coupled mechanisms. First, the modified bentonite forms a thixotropic network in ethanol. The network suspends the dense refractory particles and prevents hard settling. The card-house structure is strong at rest but breaks down under shear, so the coating remains brushable. Second, the graded bauxite and white corundum produce a dense packing structure. The coarse particles form a skeleton, and the fine particles fill the voids. This reduces the open porosity of the dried coating. A lower open porosity reduces the probability that liquid iron will penetrate the coating during pouring. It also reduces sand burn-on because the iron cannot easily reach the sand surface.
Third, the phenolic resin and polyvinyl butyral composite binder provides both high-temperature and low-temperature bonding. Polyvinyl butyral improves adhesion to the metallic chill and gives the dried film enough flexibility to resist handling damage. Phenolic resin provides high-temperature strength and a carbonaceous residue that helps keep the coating intact during the early stages of pouring. The combination is more effective than either binder alone. Fourth, the alcohol solvent evaporates quickly and ignites during the drying process. This creates a fast-drying coating that can be handled soon after application. In sand casting with external chills, fast drying is important because the coating must be applied to both the mold and the chill without delaying production.
The thermal stability of the coating can be related to the refractory aggregate and the binder residue. When the coating is heated, ethanol and residual volatiles are released first. The organic binder then decomposes and forms a carbon-rich residue. The refractory particles remain stable and provide a barrier against liquid iron. The ideal coating is one that decomposes in a controlled way and leaves a continuous refractory layer. If the binder decomposes too violently, gas pressure can disrupt the coating. If the binder content is too low, the coating may spall or lose adhesion. The optimal binder content of 2% in the A2B3C1 formulation provides enough bonding without excessive gas evolution.
10. Production Validation in External-Chill Sand Casting
I validated the optimal coating in a production environment. The casting was a heavy machine-tool bed made of HT300 gray iron. The mold was a furan resin sand mold. An external chill assembly was placed in the mold. The exposed area of a single chill was 25 cm by 15 cm, and the total coated chill area was approximately 1.2 m\(^2\). I applied two brush coats of the optimal coating. The total dry coating thickness was 0.7 mm. The coating was uniform on both the sand mold and the chill surfaces. After ignition, the coating surface cured within 5 s. The pouring temperature was 1,420 °C. After pouring and holding, the casting was cooled to 150-200 °C before shakeout.
During cleaning, the coating stripped easily from the casting surface. The chill separated from the mold without strong sticking. The casting surface was smooth, and the cleaning time was significantly reduced. The external chills were easier to recover and reuse. This is important because chill recovery reduces tooling cost and improves process consistency in sand casting. The measured surface roughness \(Ra\) was 12.5 μm. The sand adhesion area was less than 2%. There were no obvious gas pores with diameter greater than 2 mm, and there was no metal penetration or chill welding. These results satisfy the relevant requirements for gray iron castings and surface roughness comparison.
| Performance item | Before improvement | After improvement | Change |
|---|---|---|---|
| Surface roughness \(Ra\) | 25.0 μm | 12.5 μm | Reduced by 50% |
| Sand adhesion area | 8.5% | <2% | Strong reduction |
| Cleaning time | 45 min | 25 min | Reduced by 44% |
| Chill recovery rate | 65% | >90% | Improved by more than 25 percentage points |
| Gas pores \(\Phi > 2\) mm | Present | None | Eliminated |
| Metal penetration | Present | None | Eliminated |
The production results confirm that the coating is effective in real external-chill sand casting. The coating adheres to the sand mold and to the metallic chill, resists liquid iron penetration, and strips cleanly after solidification. The reduction in cleaning time is a direct economic benefit. The increase in chill recovery is also important because it reduces the need for new chills and stabilizes the chilling capacity of the process. The absence of gas pores and metal penetration indicates that the gas evolution and refractory design are well balanced.
11. Discussion of the Role of Each Component
The modified bentonite is the key to suspension stability. Without it, the dense bauxite and white corundum particles settle quickly. The coating becomes hard at the bottom and thin at the top. This causes uneven chilling and uneven surface quality in sand casting. The modified bentonite creates a yield stress network that keeps the particles suspended. The network also helps the coating adhere to vertical surfaces. When the coating is applied to a chill, it must not sag or run down. The yield stress prevents sagging at rest, while the shear-thinning behavior allows brushing. This combination is especially important for external-chill sand casting because the chill surface is smooth and non-absorbent, unlike the sand mold surface.
The graded refractory aggregate controls the packing density and the thermal barrier. The coarse fraction provides a permeable skeleton that can release gas and accommodate thermal expansion. The fine fraction fills the voids and reduces the pore size. A small pore size is beneficial because liquid iron has difficulty penetrating small pores. The white corundum increases hardness and chemical resistance. The bauxite provides good refractoriness and thermal shock resistance. The 4:1 bauxite-to-white-corundum ratio balances cost and performance. The aggregate ratio of 53:27 in the optimal formulation provides enough coarse particles for permeability and enough fine particles for packing density.
The composite binder controls adhesion, strength, and gas evolution. Polyvinyl butyral improves adhesion to the metal chill. It also increases the flexibility of the dried film. A flexible film is less likely to crack during handling or thermal expansion. Phenolic resin provides high-temperature strength and a carbon residue that helps maintain the coating during pouring. However, both binders increase viscosity and gas evolution. The optimal binder content of 2% is relatively low, but it is sufficient because the modified bentonite also contributes to green strength. The low binder content reduces gas evolution and helps avoid gas defects in sand casting.
The ethanol solvent controls drying speed and coating consistency. Ethanol evaporates rapidly and burns during ignition drying. This reduces the residual moisture in the mold and lowers the risk of gas defects. The solvent also affects the rheology of the coating. A proper solvent balance ensures that the coating wets the sand and the chill without dissolving the binder excessively or causing phase separation. The target density of \(1.60 \pm 0.02\) g/cm\(^3\) provides a reproducible way to control the solid-to-liquid ratio.
12. Interfacial Adhesion and Coating Failure Modes
I considered four possible failure modes for the coating in external-chill sand casting. The first is settling and stratification during storage. This is controlled by the modified bentonite network and the 24 h suspension rate. The second is sagging or dripping after application. This is controlled by the yield stress and the brushing grade. The third is cracking or spalling during drying and handling. This is controlled by the binder flexibility and the aggregate grading. The fourth is metal penetration or sand adhesion during pouring. This is controlled by the refractory aggregate, the coating thickness, and the high-temperature stability of the binder residue.
The coating must also accommodate the thermal expansion mismatch between the metal chill, the sand mold, and the coating. The chill expands rapidly when it is heated by the liquid iron. The sand mold expands less and can deform. The coating must remain compliant enough to avoid cracking. Polyvinyl butyral contributes to this compliance. The graded aggregate also helps because fine particles can fill gaps and reduce stress concentration. If the coating cracks, liquid iron can enter the crack and cause local penetration. The production validation showed no metal penetration, which indicates that the thermal expansion compatibility is acceptable.
The coating surface roughness after cleaning can be related to the initial coating thickness and the coating residue. If the coating is too thick, it can trap gas and create a rough surface. If it is too thin, it may not provide a continuous barrier. The optimal thickness of 0.7 mm after two brush coats is a practical compromise. It is thick enough to resist penetration and thin enough to allow rapid drying and easy stripping. The measured \(Ra\) of 12.5 μm is significantly better than the pre-improvement value of 25.0 μm. This improvement is consistent with the formation of a dense, uniform barrier layer.
13. Gas Evolution and Defect Avoidance
Gas evolution is a critical issue in alcohol-based coatings for sand casting. The gas can come from residual solvent, binder decomposition, and water or volatiles in the refractory. If the gas evolution is too high, the gas can form bubbles at the coating-metal interface. These bubbles can become pinholes or blowholes. In external-chill sand casting, the chill can act as a gas trap because it is a relatively cold, non-porous surface. The gas may not escape through the chill. It must escape through the coating and the sand mold. Therefore, the coating must have enough permeability and must not generate excessive gas.
The optimal coating has a gas evolution of 12.6 mL/g. This is well below the 20 mL/g acceptance threshold. The low binder content is the main reason for this result. The modified bentonite does not produce a large amount of gas at high temperature. The refractory aggregate is stable and does not decompose. The ethanol is removed during drying and ignition. As a result, the gas evolution is controlled. The production validation confirmed that there were no obvious gas pores with diameter greater than 2 mm. This is a strong indication that the gas evolution and coating permeability are properly balanced.
14. Suspension Stability and Storage
The 24 h suspension rate of 95% is a key achievement. In an alcohol-based coating, suspension stability is difficult because the liquid has low viscosity and low dielectric constant compared with water. The particles tend to settle, and the binder can separate. The modified bentonite solves this problem by forming a gel network in ethanol. The network is strong enough to suspend the particles for at least 24 h. This is important for industrial use because the coating may be prepared in a batch and used over a shift. If the coating settles, the operator must remix it, which adds labor and can cause inconsistent coating thickness. A high suspension rate ensures that the coating remains uniform during use.
The suspension stability also affects the brushing grade. If the coating settles, the top layer becomes thin and the bottom layer becomes thick. The thin layer may not cover the sand or the chill. The thick layer may not dry properly and may crack. The modified bentonite network prevents this stratification. It also improves the redispersibility of the coating if some settling occurs. The expansion index increase from 5.8 mL/g to 20.1 mL/g is consistent with the improved suspension behavior. The larger interlayer spacing and the grafted organic groups increase the interaction between the bentonite and the ethanol-binder system.
15. Practical Advantages in Sand Casting Production
The coating provides several practical advantages in external-chill sand casting. First, it reduces sand adhesion. Sand adhesion is a common defect when liquid iron contacts a sand mold at high temperature. The coating acts as a barrier and prevents the iron from wetting the sand. Second, it reduces metal penetration. The dense refractory layer blocks the iron from entering the coating pores. Third, it improves chill recovery. The coating prevents the iron from welding to the chill. After shakeout, the chill can be separated and reused. Fourth, it reduces cleaning time. The coating strips cleanly, leaving a smoother casting surface. Fifth, it reduces gas defects. The low gas evolution and good permeability prevent gas from being trapped at the interface.
These advantages are directly relevant to industrial sand casting. In large castings, the cost of cleaning and the cost of chill replacement can be significant. A coating that reduces cleaning time from 45 min to 25 min and increases chill recovery from 65% to more than 90% provides a clear economic benefit. The improvement in surface roughness from 25.0 μm to 12.5 μm also reduces the need for additional machining or surface treatment. The reduction in sand adhesion from 8.5% to less than 2% reduces the risk of surface defects and improves the dimensional consistency of the casting.
16. Scale-Up and Reproducibility
I prepared the coating in a wheel mixer and controlled the density to \(1.60 \pm 0.02\) g/cm\(^3\). This density control is important for reproducibility. Small changes in density can change the viscosity, the coating thickness, and the drying behavior. In production, the coating should be mixed until the refractory particles are fully wetted and the modified bentonite network is developed. The mixing time of 2.5 to 3 h is sufficient to break agglomerates and distribute the binder. A shorter mixing time may leave lumps and reduce suspension stability. A longer mixing time may entrain air and change the rheology. The coating should be stored in a sealed container to prevent ethanol evaporation. If the solvent evaporates, the density and viscosity increase, and the coating may become difficult to brush.
The orthogonal experiment and weighted normalization provide a systematic method for scale-up. The factors are defined in terms of mass ratios and mass percentages, so they can be reproduced in different batch sizes. The key responses are easy to measure with standard equipment. The optimal formulation A2B3C1 is robust because it does not rely on an extreme value of any single factor. It uses a moderate aggregate ratio, a high suspension agent content, and a low binder content. This balance is likely to be more robust in production than a formulation that is optimized for only one property.
17. Limitations and Further Development
I recognize several limitations. The current study focuses on a specific class of external-chill sand casting and a specific iron composition. Different casting temperatures, chill materials, and sand binders may require adjustments. The coating performance also depends on the surface condition of the chill. A rusty or oily chill surface may reduce adhesion. In production, the chill should be cleaned before coating. The coating thickness should be controlled because a thick coating can reduce chill efficiency. The thermal conductivity of the coating is lower than that of the metal chill, so the coating acts as a thermal barrier. The optimal thickness must balance protection and chilling. In this work, a 0.7 mm thickness provided good protection without excessive insulation.
Further development could include a more detailed study of the interfacial heat-transfer coefficient as a function of coating thickness. It could also include a study of the coating on different chill geometries and different sand casting alloys. The modified bentonite could be further optimized for different alcohol solvents or solvent blends. The binder system could be modified to reduce gas evolution even further while maintaining adhesion. The aggregate grading could be extended to include a third particle size to improve packing density. These steps would help generalize the coating for a wider range of external-chill sand casting applications.
18. Summary of Key Findings
I developed an alcohol-based coating for external-chill sand casting by combining a modified lithium bentonite suspension agent, a graded bauxite-white corundum refractory aggregate, and a phenolic resin-polyvinyl butyral composite binder. The bentonite modification used a lithium exchange step followed by organic intercalation and surface grafting. The modification increased the basal spacing from 1.26 nm to about 3.09-3.12 nm and increased the expansion index from 5.8 mL/g to 20.1 mL/g. The modified bentonite improved suspension stability and compatibility with the alcohol solvent.
The orthogonal experiment showed that the binder content has the strongest influence on condition viscosity, gas evolution, wear resistance, and brushing grade. The suspension agent content has the strongest influence on the 24 h suspension rate. The aggregate ratio affects packing density and surface quality. The weighted normalization selected A2B3C1 as the optimal formulation: a 250 mesh to 400 mesh bauxite mass ratio of 53:27, a modified suspension agent content of 5%, and a composite binder content of 2%. This formulation achieved a 24 h suspension rate of 95%, a gas evolution of 12.6 mL/g, a brushing grade of Grade I, and excellent wear resistance.
Production validation on a heavy machine-tool bed casting showed that the coating reduced surface roughness from 25.0 μm to 12.5 μm, reduced sand adhesion from 8.5% to less than 2%, reduced cleaning time from 45 min to 25 min, and increased chill recovery from 65% to more than 90%. There were no obvious gas pores with diameter greater than 2 mm and no metal penetration. The coating adhered well to both the sand mold and the external chill, dried quickly, and stripped cleanly after solidification.
I conclude that the coating is a practical solution for external-chill sand casting. It addresses the coupled problems of sand adhesion, metal penetration, gas defects, and chill recovery. The combination of a modified bentonite suspension agent, a graded refractory aggregate, and a balanced composite binder provides the required suspension stability, brushing performance, refractory resistance, and stripping behavior. The formulation is reproducible and can be scaled up for industrial sand casting. The results demonstrate that a carefully designed alcohol-based coating can improve both casting quality and production efficiency in external-chill sand casting.
| Design objective | Mechanism | Result in optimal coating |
|---|---|---|
| High suspension stability | Modified bentonite card-house network | 24 h suspension rate of 95% |
| Low gas defects | Low binder content and stable refractory | Gas evolution of 12.6 mL/g |
| Uniform application | Shear-thinning rheology and yield stress | Brushing grade of Grade I |
| Resistance to metal penetration | Graded refractory packing | No metal penetration |
| Easy cleaning and chill recovery | Balanced adhesion and stripping behavior | Cleaning time 25 min, chill recovery >90% |
| Good casting surface | Smooth, dense coating barrier | \(Ra = 12.5\) μm, sand adhesion <2% |
The final formulation and process window can be summarized as follows. The refractory aggregate consists of bauxite and white corundum at a 4:1 mass ratio. The bauxite fraction is graded at 250 mesh and 400 mesh in a 53:27 mass ratio. The total refractory mass is 100 parts. The modified suspension agent is added at 5% of the refractory mass. The composite binder is added at 2% of the refractory mass, with a phenolic resin to polyvinyl butyral mass ratio of 2.5:1. Industrial ethanol is used as the solvent. The coating density is adjusted to \(1.60 \pm 0.02\) g/cm\(^3\). The coating is mixed for 2.5-3 h and applied in two brush coats to reach a total thickness of about 0.7 mm. In external-chill sand casting, this coating provides a stable, refractory, and easily removable barrier between the chill, the sand mold, and the liquid iron.
In my assessment, the most important lesson is that external-chill sand casting requires an integrated coating design. The coating cannot be optimized for suspension alone, or adhesion alone, or gas evolution alone. The modified bentonite, graded aggregate, and composite binder must work together. When they are properly balanced, the coating improves chill efficiency, reduces cleaning cost, and raises casting quality. The results support further use of this alcohol-based coating in industrial sand casting with external chills.