As a researcher in materials engineering, I have devoted my master’s thesis to solving two critical manufacturing challenges in the production of cast steel cooling walls for blast furnaces. The first challenge is the deformation and melt-through of pre-embedded cooling water pipes during the high-temperature casting process, and the second is the severe sand foundry defect that occurs on the cooling wall surface after solidification. Through systematic numerical simulation, experimental investigation, and industrial-scale verification, I have developed effective countermeasures against both problems. This article provides a comprehensive account of my research journey, covering the numerical simulation of the casting process, the development of a solid cooling material filling method for anti-penetration, and the formulation of a novel anti-sand coating specifically designed for cast steel components. Throughout this work, the issue of sand foundry defect serves as a central theme, particularly in the context of thick-wall steel castings used in metallurgical applications.
1. Introduction and Research Background
The blast furnace remains the most important equipment in the iron and steel industry, producing more than 95% of the world’s pig iron. Over the past two decades, significant efforts have been directed toward extending the service life of blast furnaces, and the cooling wall (also known as cooling stave) has emerged as a critical component in achieving this goal. The cooling wall is installed in the furnace body, waist, bosh, and tuyere areas, where it must withstand extreme thermal loads, thermal shock, erosion by molten slag and iron, and abrasion by burden materials and gas flow. Its primary function is to dissipate heat from the furnace interior to the external environment through circulating cooling water, thereby lowering the operating temperature of refractory linings and the furnace shell.
Three main types of cooling walls are used in modern blast furnaces: cast iron, copper, and cast steel. Cast iron cooling walls suffer from low thermal conductivity and an inevitable air gap between the pipe and the base metal, which severely degrades their thermal performance. Copper cooling walls offer excellent thermal conductivity but are prohibitively expensive for widespread application. Cast steel cooling walls combine the advantages of both: they exhibit high elongation, high tensile strength, high melting point, and superior thermal shock resistance, while remaining significantly more economical than copper. The close similarity in composition between the low-alloy steel base and the cooling water pipe allows for metallurgical bonding during casting, eliminating the harmful air gap.
However, the production of cast steel cooling walls presents formidable technical barriers. The casting temperature of steel (typically 1520–1600°C) is substantially higher than that of cast iron or copper, making the pre-embedded cooling water pipe highly susceptible to deformation and melt-through. Furthermore, the large section thickness (often exceeding 250 mm) and heavy weight (above 2.5 tons) of these castings result in prolonged liquid metal solidification times, exacerbating the tendency for sand foundry defect on the casting surface. These issues have historically limited the industrial production and widespread adoption of cast steel cooling walls.
My research addresses both problems through two complementary approaches. The first is the development of an anti-penetration technique based on filling the cooling water pipe with a solid cooling material before pouring. The second is the formulation of a sinterable, easily peelable alcohol-based coating to mitigate sand foundry defect. This work aims to answer the following research questions:
- Can numerical simulation accurately predict the temperature field of cooling water pipes during mold filling and solidification?
- What is the optimal composition and process parameter for the solid cooling material used to fill the cooling pipe?
- Can a metallurgical bond be achieved between the cooling pipe and the base metal while preventing melt-through?
- How can an alcohol-based coating be formulated to minimize sand foundry defect on thick-wall steel castings?
2. Numerical Simulation Analysis of the Casting Process
Numerical simulation using the finite element method provides a powerful tool for understanding and optimizing the casting process. In this study, I utilized the ProCAST software package to simulate both mold filling and solidification of the cast steel cooling wall. The simulation workflow is illustrated in the following steps:
- Three-dimensional CAD modeling of the cooling wall base, cooling water pipe, and pouring system using UG software
- Mesh generation using MeshCAST, resulting in 269,626 triangular surface elements and 5,433,693 tetrahedral volume elements
- Material property assignment for ZG35CrMnSi low-alloy steel (base) and 20G steel (cooling pipe)
- Boundary condition definition, including heat transfer coefficients between different interfaces
- Solution of the governing equations for fluid flow and heat transfer
2.1 Gating System Design
Based on the structural characteristics of the cooling wall, I designed a bottom-gating system to achieve smooth mold filling and prevent slag entrapment. The minimum cross-sectional area of the gating system was calculated using the following formula:
$$S_{\text{阻}} = \frac{m}{\mu \cdot \tau \cdot \rho \cdot H_{\text{均}}}$$
where $S_{\text{阻}}$ is the minimum cross-sectional area (cm²), $m$ is the mass of liquid metal (kg), $\mu$ is the flow coefficient, $\tau$ is the filling time (s), $\rho$ is the density of liquid metal (g/cm³), and $H_{\text{均}}$ is the average static pressure head (cm).
The design parameters and calculated results are summarized in Table 1.
| Parameter | Value | Unit |
|---|---|---|
| Mass of casting | 2686.4 | kg |
| Flow coefficient $\mu$ | 0.32 | — |
| Filling time $\tau$ | 32.4 | s |
| Density $\rho$ | 7.9 | g/cm³ |
| Average static head $H_{\text{均}}$ | 1185 | mm |
| Minimum cross-sectional area $S_{\text{阻}}$ | 1256.63 | cm² |
| Sprue diameter | 60 | mm |
| Runner diameter | 70 | mm |
| Riser diameter | 40 | mm |
2.2 Mold Filling Simulation Results
The mold filling simulation was performed with a pouring temperature of 1550°C, an initial mold temperature of 25°C, and a pouring rate of 30 kg/s under normal gravity. The simulation results revealed three distinct stages during mold filling:
At t = 13.4 s, the molten metal had nearly filled the bottom section of the casting cavity. At t = 63.6 s, the molten metal made contact with the cooling water pipe and began to envelop it. By t = 97.7 s, the cavity was completely filled, and the metal had entered the riser system. The total filling time was approximately 100.8 s.
From a thermal perspective, the cooling water pipe temperature remained largely unchanged until the molten metal made contact. At t = 63.6 s, the lower surface of the pipe experienced a rapid temperature rise to above 620°C. By the end of filling, the maximum temperature of the pipe remained below 1100°C, which is still far lower than the solidus temperature of the pipe material (1473.2°C). Therefore, the cooling water pipe does not suffer melt-through during the mold filling stage.
2.3 Solidification Simulation Results
The solidification simulation revealed a more concerning picture. I monitored the temperature distribution at six cross-sectional positions (A through F) along the cooling water pipe, with positions B, C, D, and E located closest to the ingates and position E adjacent to the thickest section of the casting. The simulation results are summarized in Table 2.
| Section | Maximum Temperature (°C) | Risk of Melt-Through |
|---|---|---|
| A | < 1490 | Low |
| B | < 1495 | Low |
| C | < 1498 | Low |
| D | < 1500 | Moderate |
| E | 1502–1510 | High |
| F | < 1480 | Low (poor bonding) |
The results clearly show that section E, located at the maximum wall thickness position (430 mm), experiences the highest temperatures during solidification. The local maximum temperature at the lower surface of the pipe reached 1504.1°C, which exceeds the liquidus temperature of the pipe material (1503.6°C). This finding confirms that the cooling water pipe is at significant risk of melt-through during the solidification stage, specifically at the lower surface adjacent to section E.
2.4 Effect of Pouring Temperature on Interface Fusion
I systematically investigated the influence of pouring temperature on the metallurgical bonding between the cooling pipe and the base metal. Three pouring temperatures were evaluated: 1520°C, 1550°C, and 1580°C, while maintaining all other parameters constant. The results are presented in Figure 1, which plots the proportion of the pipe surface achieving metallurgical bonding as a function of pouring temperature.
The relationship between pouring temperature and the fraction of metallurgical bonding can be expressed as:
$$K(T) = \frac{S – S_1(T) – S_2(T)}{S}$$
where $K$ is the fraction of the pipe surface achieving metallurgical bonding, $S$ is the total embedded surface area of the pipe, $S_1$ is the area that melts through, and $S_2$ is the area that fails to reach the solidus temperature.
Table 3 summarizes the simulated bonding fractions at different pouring temperatures.
| Pouring Temperature (°C) | Melt-Through Area (%) | Non-Bonding Area (%) | Metallurgical Bonding Fraction (%) |
|---|---|---|---|
| 1520 | 0 | 15.5 | 84.5 |
| 1550 | 4.3 | 4.5 | 91.2 |
| 1580 | 12.5 | 1.2 | 86.3 |
At 1520°C, the metal temperature was insufficient to fully melt the pipe surface, leaving approximately 15.5% of the pipe without metallurgical bonding. At 1580°C, excessive superheat caused 12.5% of the pipe to melt through. The optimal pouring temperature was identified as 1550°C, yielding a maximum bonding fraction of 91.2%. This temperature provides sufficient heat for bonding while avoiding excessive melt-through.
2.5 Simulation of Solid Cooling Material Filling
Based on the identified risk of melt-through at section E, I proposed a preventive measure: filling the interior of the cooling water pipe with a solid cooling material before pouring. The mechanism of this approach is that the solid material absorbs heat from the pipe during mold filling and solidification, providing a chilling effect on the surrounding molten metal. This reduces the pipe temperature and prevents it from melting through while still allowing sufficient fusion for metallurgical bonding.
Numerical simulation was performed with a heat transfer coefficient of 800 W/(m²·K) between the cooling pipe and the solid filling material. The simulated temperature distribution is shown in Table 4.
| Section | Maximum Temperature (°C) | Minimum Temperature (°C) | Melt-Through Risk | Bonding Capability |
|---|---|---|---|---|
| A | 1485.3 | 1478.1 | None | Good |
| B | 1488.2 | 1477.5 | None | Good |
| C | 1489.1 | 1476.9 | None | Good |
| D | 1490.8 | 1476.6 | None | Good |
| E | 1492.4 | 1476.4 | None | Good |
| F | 1481.3 | 1476.2 | None | Good |
These simulation results demonstrate that filling the cooling pipe with a solid cooling material can effectively reduce the maximum temperature from 1504.1°C to 1492.4°C — well below the melting point of the pipe material (1503.6°C) — while maintaining the minimum temperature above the solidus temperature (1473.2°C). Therefore, the pipe would neither melt through nor fail to bond. The simulation validated the feasibility of this anti-penetration approach.
3. Experimental Study of Resin Sand Filling for Anti-Penetration
Based on the numerical simulation results, I proceeded to experimentally investigate the use of resin sand as the solid cooling material. The selection of the sand type was based on the following criteria:
- High heat storage capacity for effective chilling
- Excellent stability at high temperatures
- No chemical reaction with the pipe material
- Good flowability for filling and easy removal after casting
Three candidate sands were evaluated: ceramsite (宝珠砂), zircon sand, and chromite sand. Their thermophysical properties, which are relevant to their performance as cooling materials, are presented in Table 5.
| Sand Type | Refractoriness (°C) | Bulk Density (t/m³) | Thermal Conductivity (W/m·K) | Linear Expansion (10⁻⁶K⁻¹) | Specific Heat (J/kg·K) | Angular Coefficient |
|---|---|---|---|---|---|---|
| Ceramsite | 1800 | 1.9–2.0 | 0.5–0.6 | 6 | 2210 | 1.05 |
| Zircon sand | 1830 | 2.93 | 0.8–0.9 | 4.6 | 628 | 1.50 |
| Chromite sand | 1830 | 2.8 | 0.65 | 8 | 1214 | 1.35 |
3.1 Effect of Sand Type on Filling Efficiency
The experimental procedure involved mixing each sand with a phenolic resin binder (1% by weight) and a curing agent (40% of the binder weight). The prepared resin sand was then compressed into a scale-model cooling pipe (1:5 scale) using compressed air at 0.3 MPa. The filling efficiency was calculated using the following formula:
$$K_{\text{filling}} = \frac{m_1 – m_0}{\rho \cdot V}$$
where $K_{\text{filling}}$ is the filling efficiency, $m_0$ is the pipe mass, $m_1$ is the total mass of pipe and filled sand, $\rho$ is the density of the packed sand, and $V$ is the internal volume of the pipe.
The experimental results are presented in Table 6.
| Sand Type | Filling Efficiency (%) |
|---|---|
| Ceramsite | 87.4 |
| Chromite sand | 73.2 |
| Zircon sand | 62.1 |
Ceramsite exhibited the highest filling efficiency due to its nearly spherical particle shape (angular coefficient of 1.05), which reduces inter-particle friction and improves flowability. Zircon sand, with angular particles (coefficient of 1.50), showed the lowest filling efficiency. Based on this result, ceramsite was selected as the preferred base sand.
3.2 Effect of Filling Pressure
Using ceramsite as the base sand with 1% binder, I varied the compressed air pressure from 0.1 to 0.9 MPa. Table 7 shows the results.
| Filling Pressure (MPa) | Filling Efficiency (%) |
|---|---|
| 0.1 | 75.2 |
| 0.3 | 87.4 |
| 0.5 | 96.5 |
| 0.7 | 97.1 |
| 0.9 | 93.8 |
The filling efficiency initially increased with pressure, peaking at 0.7 MPa, then decreased. The decrease at 0.9 MPa can be attributed to excessive pressure causing sand compaction at horizontal sections of the pipe and creating a pressure peak that impeded further sand movement. As the efficiency at 0.5 MPa was only slightly lower than at 0.7 MPa, the lower pressure was selected for practical reasons, as it places less demand on equipment.
3.3 Effect of Binder Content
I then varied the binder content from 0.8% to 1.6% while maintaining the filling pressure at 0.5 MPa. Table 8 summarizes the results.
| Binder Content (%) | Filling Efficiency (%) |
|---|---|
| 0.8 | 88.3 |
| 1.0 | 94.7 |
| 1.2 | 99.1 |
| 1.4 | 95.2 |
| 1.6 | 91.8 |
The optimum binder content was 1.2%, achieving a filling efficiency of 99.1%, which indicates essentially complete filling. Increasing the binder content beyond 1.2% caused the sand particles to agglomerate into larger clusters that hindered flow.
3.4 Sand Removal Study
To simulate the service conditions of the cooling pipe during casting, filled pipes were heated to 1495°C for 35 minutes and then cooled in air. The resin sand was removed by vibration and impact. The removal efficiency was calculated as follows:
$$K_{\text{removal}} = \frac{m_4}{m_3 – m_0}$$
where $m_3$ is the mass of the pipe with sand before cleaning, $m_4$ is the mass of sand removed, and $m_0$ is the mass of the empty pipe.
Initial tests showed that only about 60% of the sand could be removed because the high temperature caused partial sintering between sand particles, forming large agglomerates that could not flow out of the pipe. To improve removal efficiency, I added varying amounts of corundum sand (which has excellent high-temperature stability, with a refractoriness above 2050°C) to the ceramsite base sand. The results are summarized in Table 9.
| Corundum Content (% of total sand weight) | Removal Efficiency (%) |
|---|---|
| 5 | 72.5 |
| 10 | 86.3 |
| 15 | 96.8 |
| 20 | 97.5 |
| 25 | 98.1 |
Adding 15% corundum sand to the ceramsite base sand increased the removal efficiency to 96.8%, which satisfies industrial requirements. The optimal formulation was determined as: ceramsite (85%) + corundum sand (15%) + phenolic resin (1.2% of total sand weight) + curing agent (40% of resin weight).
4. Industrial-Scale Casting Verification
To validate the laboratory findings, I conducted an industrial-scale trial in a production facility. The cooling water pipe was filled with the optimized resin sand formulation using a filling pressure of 0.5 MPa. The pipe was then secured in the mold using steel rods and spacers to prevent displacement during pouring. The mold was prepared using sodium silicate sand with a white corundum alcohol-based coating applied to the cavity surface.
Following the pouring at 1550°C and subsequent cooling, the casting was sectioned along the pipe axis. Macroscopic examination revealed excellent bonding between the cooling pipe and the base metal, with no visible gaps, deformation, or melt-through at any location along the pipe length.
4.1 Metallographic Analysis
Samples were extracted from the bonded interface region and prepared for metallographic examination. The interface microstructure was observed using optical microscopy and scanning electron microscopy. The key observations were as follows:
The microstructure of both the cooling wall base and the cooling pipe consisted of pearlite and ferrite phases, with the bright regions being ferrite and the darker regions being pearlite. A slight difference in the ferrite-to-pearlite ratio was noted, attributed to the different carbon contents of the two materials (the 20G pipe contains approximately 0.2% carbon, slightly higher than that of the ZG35CrMnSi base). The interface zone displayed the same phases, confirming that metallurgical reaction occurred between the base metal and the pipe.
Scanning electron microscopy with energy-dispersive X-ray spectroscopy (EDS) was used to analyze the elemental diffusion profile across the interface. The concentrations of Si, Cr, and Mn, which are present in higher amounts in the base metal, showed a decreasing trend across the interface, confirming that these elements diffused from the base metal into the pipe surface and formed solid solutions with iron.
4.2 Mechanical Testing
Shear testing was conducted on specimens extracted from the interface region to quantify the bonding strength. The average shear strength was determined to be 318 MPa. The fracture surface morphology revealed characteristic shear dimples, indicating ductile fracture rather than brittle failure. The presence of parabolic and semi-elliptical dimples oriented along the shear direction confirmed that the interface possessed high strength and that true metallurgical bonding had been achieved.
5. Development of Anti-Sand Coating for Cast Steel Cooling Walls
After addressing the melt-through problem, I turned my attention to the second critical issue: the sand foundry defect on the surface of cast steel cooling walls. The thick sections and prolonged solidification times of these castings make them particularly susceptible to this defect. I developed a sinterable, easily peelable alcohol-based coating designed specifically for thick-wall steel castings.
5.1 Coating Composition and Preparation
The coating formulation involved the following components:
- Refractory aggregate: White corundum (Al₂O₃ ≥ 98.5%), selected for its high refractoriness (2000–2200°C) and low thermal expansion
- Flux agents: Potassium feldspar powder and iron oxide powder to promote sintering at casting temperatures
- Suspending agents: Lithium bentonite (pre-swollen in water at 1:2 ratio) and polyvinyl butyral (PVB)
- Binders: Phenolic resin (2123) with hexamine as a cross-linking agent
- Surfactant: Tween-80 to improve wetting and penetration of the sand mold
- Carrier: Ethanol (≥95% purity)
Table 10 provides the complete formulation that I developed.
| Component | Amount (g) | Percentage of Total Powder (%) |
|---|---|---|
| White corundum | 100 | 73.5 |
| Potassium feldspar | 24 | 17.6 |
| Iron oxide | 12 | 8.8 |
| Swollen lithium bentonite | 13.3 | 9.8 |
| Phenolic resin (2123) | 2.5 | 1.8 |
| Hexamine | 0.5 | 0.36 |
| PVB | 0.16 | 0.12 |
| Tween-80 | 0.1 | — |
| Ethanol | 90 | 66 |
The preparation procedure involved sequential mixing of components: first the swollen bentonite was dispersed in ethanol, followed by the phenolic resin and hexamine, then PVB and Tween-80, and finally the feldspar, iron oxide, and white corundum powders. The mixture was stirred for 10–15 minutes to ensure homogeneity.
5.2 Effect of Lithium Bentonite Content on Coating Properties
In the first series of experiments, I varied the swollen lithium bentonite content from 4 g to 16 g while keeping all other components constant (white corundum 100 g, potassium feldspar 24 g, iron oxide 12 g, phenolic resin 2.5 g, hexamine 0.5 g, PVB 0.3 g, Tween-80 0.1 g, ethanol 90 g). The measured properties are shown in Table 11.
| Bentonite (g) | Viscosity (s) | Density (g/cm³) | 2h Suspension (%) | 24h Suspension (%) |
|---|---|---|---|---|
| 4 | 15.96 | 1.50 | 84 | 73.5 |
| 8 | 16.26 | 1.51 | 89 | 74 |
| 12 | 16.66 | 1.49 | 96 | 79 |
| 16 | 20.25 | 1.56 | 97 | 83 |
As the bentonite content increased, both viscosity and suspension stability improved. The bentonite particles form a three-dimensional network structure that keeps the refractory aggregate suspended in the alcohol carrier. However, excessive bentonite increased the viscosity beyond practical levels for brushing. A content of 12 g was selected as optimal.
5.3 Effect of Phenolic Resin Content
Next, I varied the phenolic resin content from 2 g to 3.5 g while keeping the bentonite at 12 g. Table 12 presents the results.
| Resin (g) | Viscosity (s) | Density (g/cm³) | 2h Suspension (%) | 24h Suspension (%) |
|---|---|---|---|---|
| 2.0 | 15.20 | 1.47 | 92 | 79 |
| 2.5 | 16.66 | 1.49 | 96 | 83.5 |
| 3.0 | 22.22 | 1.48 | 93.5 | 79 |
| 3.5 | 21.09 | 1.49 | 93 | 78.5 |
The resin binder improves the bonding strength of the dried coating but also increases viscosity. The optimum content was 2.5 g, which balanced good suspension with acceptable viscosity.
5.4 Orthogonal Experiment Optimization
To account for potential interactions between components, I conducted an L₁₆(4⁴) orthogonal experiment with four factors at four levels each: lithium bentonite (A), phenolic resin (B), PVB (C), and ethanol (D). The results were analyzed using range analysis to identify the optimal combination.
For viscosity, the order of influence was D > C > A > B, with the optimal combination being A₃B₂C₁D₁. For 2-hour suspension, the order was D > A > C > B, with the optimal combination being A₄B₃C₁D₄. For 24-hour suspension, the order was D > A > B > C, with the optimal combination being A₄B₁C₄D₄.
The comprehensive analysis yielded the final optimized formulation as presented in Table 10.
5.5 Effect of Flux Ratio on High-Temperature Performance
The flux system (potassium feldspar and iron oxide) is critical for controlling the sintering behavior of the coating at casting temperatures. I evaluated seven different ratios of potassium feldspar to iron oxide, keeping the total flux content at 36 g per 100 g of white corundum. The thermal shock resistance of the coating was assessed by heating coated cylindrical sand specimens to 1300°C for 2.5 minutes and observing crack formation and coating detachment. Table 13 summarizes the results.
| Flux Ratio (Feldspar:Iron Oxide) | Thermal Shock Rating |
|---|---|
| 0:1 | IV |
| 1:3 | III |
| 1:2 | II |
| 1:1 | II |
| 2:1 | I |
| 3:1 | II |
| 1:0 | III |
| No flux | IV |
The thermal shock ratings follow the standard classification: Grade I indicates no cracks or very fine cracks, Grade II indicates a network of fine cracks less than 0.5 mm wide, Grade III indicates dendritic cracks 0.5–1 mm wide, and Grade IV indicates severe cracking with widths greater than 1 mm.
Pure white corundum exhibits poor sintering behavior at 1300°C and cracks under thermal shock. The addition of flux promotes liquid-phase sintering, which improves the integrity of the coating layer. A ratio of 2:1 (feldspar to iron oxide) produced the best result, achieving Grade I. The potassium oxide, silicon dioxide, and iron oxide in these flux materials promote the sintering of aluminum oxide, creating a dense ceramic layer that prevents metal penetration and reduces sand foundry defect.
5.6 Industrial Trial of the Developed Coating
The optimized coating was applied to the sand mold surface for the production of actual cast steel cooling walls. The application procedure involved two steps: first, the coating was sprayed onto the mold surface to achieve an even layer; second, a soft brush was used to level the coating in the grooves and recesses where spraying tended to accumulate excess material. The coated mold was then ignited to evaporate the ethanol and solidify the coating.
After casting and subsequent shakeout, the cooling wall surface was examined. Table 14 summarizes the observed conditions.
| Surface Location | Condition After Shakeout | Remarks |
|---|---|---|
| Flat surface (front) | Clean, sand mold detached completely | No sand foundry defect |
| Flat surface (side) | Clean, sand mold detached completely | No sand foundry defect |
| Groove positions | Spot-like adhesion observed | Easily removed after impact |
More than 95% of the cooling wall surface area required no additional cleaning after the initial shakeout. The spot-like adhesion in the grooves was of low strength and could be removed by light impact. This level of performance satisfies the production requirement and demonstrates the effectiveness of the developed coating in mitigating sand foundry defect on thick-wall steel castings.
6. Conclusions and Outlook
Through this comprehensive study, I have arrived at the following conclusions:
- Numerical simulation of the casting process: The mold filling simulation revealed that the cooling water pipe is safe during mold filling, with a maximum temperature below 1100°C. However, during solidification, the pipe reaches local temperatures of 1502–1510°C, exceeding or approaching its melting point of 1503.6°C. The critical melt-through location is at the lower surface of the pipe in the thickest section of the casting. The optimal pouring temperature is 1550°C, which yields a metallurgical bonding fraction of 91.2%.
- Anti-penetration using solid cooling material: Filling the cooling water pipe with a solid cooling material is a practical and effective anti-penetration measure. The optimal formulation was determined as ceramsite sand (85%) mixed with corundum sand (15%), using 1.2% phenolic resin binder and a filling pressure of 0.5 MPa. This formulation achieves 99.1% filling efficiency and 96.8% removal efficiency.
- Metallurgical bonding: Industrial-scale verification confirmed that the cooling pipe and base metal achieve true metallurgical bonding, with an average shear strength of 318 MPa. The interface microstructure consists of pearlite and ferrite, and elemental diffusion (Si, Cr, Mn) was confirmed across the interface.
- Anti-sand coating: The developed alcohol-based coating, using white corundum as the aggregate and a 2:1 ratio of potassium feldspar to iron oxide as flux, provides excellent thermal shock resistance (Grade I) and effectively prevents sand foundry defect on cast steel cooling walls. The coating allows for easy cleaning of more than 95% of the casting surface area.
While this research has successfully addressed the two main challenges in cast steel cooling wall production, several aspects warrant further investigation:
- Combined anti-penetration strategies: Exploring the combined use of internal solid cooling materials and external protective measures (such as chills or insulating coatings) to provide additional safety margins during casting.
- Water-soluble solid cooling materials: Developing water-soluble materials that can be removed by simple flushing with water after casting, eliminating the need for mechanical cleaning.
- Advanced flux systems: Investigating alternative flux agents and their concentrations to further improve the sintering behavior and high-temperature performance of the coating.
- Modeling of coating sintering: Developing predictive models for coating sintering behavior to accelerate the formulation optimization process.
In conclusion, my research has contributed to the understanding and practical resolution of two critical manufacturing challenges in cast steel cooling wall production. The numerical simulation framework, the anti-penetration technique, and the anti-sand coating developed in this work collectively represent significant progress toward the reliable and economical production of high-quality cast steel cooling walls. The findings directly support the broader application of cast steel cooling walls in blast furnaces, which will contribute to improved furnace longevity, reduced maintenance costs, and enhanced efficiency in the iron and steel industry. My work emphasizes the importance of combining computational simulation with experimental validation and industrial-scale verification to develop robust manufacturing solutions. The issue of sand foundry defect, which initially appeared to be a downstream problem, was ultimately solved through systematic materials engineering and proper formulation development, demonstrating the interconnected nature of casting process challenges.