In my extensive work with resin sand casting, particularly in the production of thin-walled components like cylinder blocks and heads for diesel engines, I have consistently faced the challenge of mechanical burn-on. This defect, where metal penetrates into the sand mold or core, severely compromises surface quality and necessitates costly cleaning operations. Through rigorous analysis and experimentation, I have delved into the root causes and developed effective countermeasures. This article shares my firsthand insights into the mechanisms behind burn-on in cold-set furan resin sand casting and outlines practical solutions that have proven successful in industrial applications.
Resin sand casting, utilizing cold-set furan resin binders, offers numerous advantages such as excellent dimensional accuracy, good collapsibility, and high productivity. However, its susceptibility to mechanical burn-on, especially in intricate cores with thin sections, remains a significant drawback. The phenomenon is not merely a superficial issue but stems from complex interactions between the sand core, coating, and molten metal during pouring and solidification. My investigation began with characterizing the high-temperature behavior of the resin sand system.
The core of the problem lies in the degradation of the resin binder under thermal load. In a typical resin sand casting process, the sand mix consists of silica sand, a furan resin binder (usually around 1.0-1.5% by weight), and a acid catalyst. At room temperature, this provides adequate strength for handling. However, upon exposure to molten iron at temperatures exceeding 1350°C, the organic resin undergoes rapid pyrolysis. To quantify this, I conducted high-temperature strength tests. Specimens with dimensions of φ50 mm × 50 mm were heated to 1000°C and held for a short duration, simulating the thermal shock during casting. The results were stark: the surface resin binder completely decomposed and burned away, leaving the sand grains loosely bonded. When a load of merely 0.05 MPa was applied under these conditions, the core crumbled. This indicates that the core loses its structural integrity well before the metal solidifies, creating pathways for penetration.
The high-temperature strength decay can be modeled using an exponential decay function, reflecting the thermal decomposition kinetics of the resin:
$$ \sigma(T, t) = \sigma_0 \cdot e^{-k(T) \cdot t} $$
Where $\sigma(T, t)$ is the strength at temperature $T$ and time $t$, $\sigma_0$ is the initial room-temperature strength, and $k(T)$ is a temperature-dependent degradation rate constant. For furan resin, $k(T)$ increases dramatically above 500°C. This loss of cohesive strength between sand grains is the first prerequisite for mechanical burn-on.
Furthermore, the thermal expansion of silica sand exacerbates the issue. Silica undergoes phase transformations upon heating, notably a rapid expansion near 573°C during the α- to β-quartz transition. The linear thermal expansion can be approximated by:
$$ \frac{\Delta L}{L_0} = \alpha \cdot \Delta T + \beta \cdot (\Delta T)^2 $$
Where $\alpha$ and $\beta$ are coefficients specific to the sand. This expansion induces tensile stresses in the core surface and, more critically, in any applied coating. If the coating cannot accommodate this strain, it cracks, providing direct conduits for metal ingress.
The role of the coating is therefore paramount. Initially, we used a double-layer coating with a graphite-based wash. Graphite has high thermal conductivity ($\lambda \approx 100-150 \text{ W/m·K}$), which, while beneficial for certain aspects, proved detrimental in this context. The high conductivity caused rapid heat transfer into the sand core, accelerating its thermal expansion and the resin’s degradation. Moreover, during pouring, the graphite and organic binders in the coating oxidize and burn, weakening the coating’s sintered structure. The coating’s failure stress under thermal load can be expressed as:
$$ \sigma_c = \sigma_{c0} – \gamma \cdot \dot{T} $$
Where $\sigma_{c0}$ is the intrinsic coating strength, $\gamma$ is a thermal shock sensitivity parameter, and $\dot{T}$ is the heating rate. For graphite-based coatings, $\gamma$ is high, leading to premature cracking.
The mechanism of metal penetration into the porous sand core can be described by modified Darcy’s law for flow in a deformable medium. The penetration velocity $v$ is given by:
$$ v = \frac{K}{\mu \cdot \phi} \cdot \frac{dP}{dx} $$
Where $K$ is the permeability of the sand core (which increases as resin degrades and grains rearrange), $\mu$ is the dynamic viscosity of the molten metal, $\phi$ is the porosity, and $dP/dx$ is the pressure gradient driven by metallostatic head and capillary forces. The critical pressure for penetration, $P_{crit}$, is related to the effective pore radius $r$ and the metal-sand contact angle $\theta$:
$$ P_{crit} = \frac{2 \gamma_{lv} \cos \theta}{r} $$
Where $\gamma_{lv}$ is the liquid-vapor surface tension of the metal. As the resin binder disappears, sand grains shift, enlarging pores (increasing $r$), which drastically reduces $P_{crit}$ and allows metal to seep in deeply even under modest pressure.
Having identified these interconnected mechanisms, I focused on two synergistic strategies to combat burn-on in resin sand casting: modifying the core sand composition and redesigning the coating system. The first approach involved incorporating iron oxide (Fe₂O₃) powder into the resin sand mix. We experimented with additions ranging from 0.5% to 3% by weight. The optimal balance was found at 1.5-2.0% addition. This modification served multiple purposes:
1. Pore Filling and Glassy Phase Formation: At high temperatures, iron oxide reacts with silica sand to form low-melting-point ferrosilicate phases, which flow and seal surface pores. The reaction can be simplified as:
$$ \text{Fe}_2\text{O}_3 + \text{SiO}_2 \xrightarrow{\text{high } T} \text{Fe-silicate glass} $$
This glassy layer acts as a physical barrier against metal penetration.
2. Reduction of Thermal Expansion: The iron oxide alters the thermal expansion characteristics of the sand mixture. Measured data showed a reduction in the linear expansion coefficient by approximately 15-20% in the critical 500-800°C range, mitigating coating stress.
3. Generation of Protective Atmosphere: Iron oxide decomposes and reacts with carbon from the decomposed resin, producing a reducing gas atmosphere (containing CO, H₂) at the metal-mold interface, which can improve surface finish by minimizing oxidation.
4. Enhanced High-Temperature Cohesion: The formed glassy phase provides a viscous, flexible bond between sand grains after resin burnout, maintaining some integrity under heat, unlike the brittle carbonaceous residue from pure resin.
The mixing procedure was adjusted as follows: first, dry-mixing the base sand and iron oxide powder, then adding the liquid catalyst and mixing thoroughly, followed by adding the furan resin and mixing until homogeneous before discharge. The properties of the modified sand are summarized in Table 1.
| Property | Standard Resin Sand | Resin Sand + 2% Fe₂O₃ | Test Method |
|---|---|---|---|
| Tensile Strength (24h, MPa) | 1.2 – 1.4 | 1.1 – 1.3 | Standard specimen |
| High-Temp Strength (1000°C, MPa)* | ~0.02 | ~0.08 | Hot compression test |
| Linear Thermal Expansion (0-800°C, %) | 1.45 | 1.18 | Dilatometry |
| Permeability | ~180 | ~175 | Standard permeability test |
| Burn-On Tendency Index** | High (8-9) | Low (2-3) | Practical casting test |
* Approximate value under rapid heating. ** Scale 1-10, 1 being minimal burn-on.
The second strategy involved developing a novel double-layer coating system to replace the graphite-based one. The requirements were: low thermal conductivity to shield the core, good sinterability to form a dense barrier, and adequate permeability to allow gas escape. The new system comprised:
Primary (Bottom) Layer: A water-based refractory wash with high filling power and low thermal conductivity.
- Composition: 100 parts quartz flour (SiO₂, -200 mesh), 4 parts bentonite, 2 parts sodium silicate (as % of bentonite), and 1.5 parts dextrin.
- Rationale: Quartz flour has a low thermal conductivity ($\lambda \approx 1-2 \text{ W/m·K}$), providing excellent thermal insulation. Its fine particles effectively block sand grain interstices. This layer’s primary function is to protect the core’s integrity during the initial thermal shock.
Secondary (Top) Layer: A fast-drying, sinterable coating designed to form a ceramic-like barrier at casting temperatures.
- Composition: 40 parts graphite powder, 30 parts quartz flour, 15 parts talc, 10 parts iron oxide powder, 3 parts bentonite, 1.5 parts sodium silicate (as % of bentonite), and 2 parts dextrin.
- Rationale: This multi-component system (C-SiO₂-MgO-Fe₂O₃) is designed to undergo sintering and liquid phase formation at casting temperatures, creating a continuous, impervious vitrified layer. The presence of iron oxide promotes the formation of a stable, high-melting-point interface.
The sintering behavior of the top layer can be described by a viscous flow model. The densification rate is proportional to the viscosity $\eta$ and the driving force from surface energy $\gamma_s$:
$$ \frac{d\rho}{dt} = \frac{A \cdot \gamma_s}{r \cdot \eta(T)} $$
Where $\rho$ is the density, $A$ is a geometric constant, $r$ is particle radius, and $\eta(T)$ decreases with temperature according to an Arrhenius law: $\eta(T) = \eta_0 \exp(E_\eta / RT)$. The designed composition ensures $\eta(T)$ drops sufficiently around 1200°C to allow rapid sealing.
To evaluate the coating performance, I conducted a series of tests, results of which are in Table 2. The cracking time was assessed by coating standard resin sand cylinders and subjecting them to a propane torch flame, simulating sudden heating.
| Coating System | Thermal Conductivity (W/m·K, est.) | Cracking Time at 1100°C (seconds) | Post-Heating Integrity | Metal Penetration Depth (mm)* |
|---|---|---|---|---|
| Old Graphite-Based Double Layer | ~80 | 8-12 | Severe cracking, spalling | 1.5 – 2.5 |
| New Double Layer (Quartz + Sinterable) | ~5-10 | 25-35 | Intact, sintered surface | 0.1 – 0.3 |
* Measured on cast test plates of 6mm thickness.
The combined application of iron oxide-doped resin sand and the new double-layer coating was implemented in the production of cylinder heads for a 135-series diesel engine. These castings are particularly challenging due to their complex internal water jacket cores with wall thicknesses often below 5mm. The production trial involved several hundred castings. The results were quantitatively assessed based on the need for post-casting cleaning (grinding) and the visual quality of the as-cast surface in the internal passages. A summary is provided in Table 3.
| Process Version | Burn-On Incidence Rate (%) | Average Cleaning Time per Casting (minutes) | Surface Roughness (Ra, μm) in Critical Areas | Overall Yield Improvement (%) |
|---|---|---|---|---|
| Original Process (Standard sand, Graphite coating) | ~85 | 25-30 | 25-50 | Baseline |
| Modified Process (Fe₂O₃ sand, New coating) | < 10 | 3-5 | 12-20 | ~12 |
The improvement was profound. The internal surfaces of the cylinder heads, previously plagued by extensive metal penetration requiring arduous de-coring and grinding, now exhibited a clean, readily shake-out sand core and a smooth metal surface. The mechanistic success can be attributed to the following sequence: the low-conductivity primary coating delayed the heat transfer to the core, reducing the initial thermal shock. As temperature rose, the iron oxide in the sand began to react, forming a sealing layer and providing a more ductile grain bond. The top coating sintered into a dense, continuous shield, blocking any potential metal pathways. The reduced thermal expansion of the modified sand core further prevented cohesive failure and coating fracture.
In conclusion, my investigation into the burn-on defect in cold-set furan resin sand casting revealed that it is a consequence of the synergistic failure of the sand core’s high-temperature strength and the protective coating’s integrity under thermal stress. The key to prevention lies in proactively managing these failure modes. By incorporating a small percentage of iron oxide powder into the resin sand mix, we fundamentally improve the core’s behavior at elevated temperatures, promoting pore sealing and reducing thermal stress. Complementing this with a rationally designed double-layer coating system—featuring a thermally insulating base and a sinterable top layer—creates a robust barrier against metal penetration. This integrated approach has transformed the viability of resin sand casting for producing high-quality, thin-walled castings, significantly reducing rework and enhancing surface finish. Future work in resin sand casting could explore the optimization of iron oxide particle size and distribution for even better performance, as well as the development of more sustainable coating materials, ensuring that resin sand casting remains a competitive and reliable manufacturing process.