In my research, I investigated the thermophysical parameters of molding sand used in sand casting processes. The accuracy of these parameters is critical for simulating the solidification and cooling of sand casting parts, which directly impacts production quality. Through numerical simulation and experimental methods, I aimed to determine how thermal conductivity and specific heat vary with temperature, and how these properties influence the manufacturing of sand casting parts. This study provides essential data for improving the precision of casting simulations, thereby enhancing the reliability and efficiency of producing sand casting parts.
Sand casting is a widely used manufacturing method for creating metal components, including complex sand casting parts. The process involves pouring molten metal into a mold made of molding sand, which then solidifies. The thermal behavior of the molding sand during this phase determines the cooling rate, microstructure, and mechanical properties of the final sand casting parts. However, molding sand is a porous, non-homogeneous material with variable thermophysical properties due to factors like binder composition, coating, and compaction density. This variability makes it challenging to use literature values for accurate simulations, often leading to errors in predicting temperature distributions and solidification patterns in sand casting parts. Therefore, I conducted this study to experimentally derive thermophysical parameters tailored to specific production conditions, ensuring better guidance for actual foundry operations.
Previous studies have highlighted the importance of accurate thermophysical data for casting simulations. For instance, researchers have noted that discrepancies in sand properties can result in significant deviations in predicted cooling curves, affecting the quality of sand casting parts. In my work, I built upon these insights by combining direct measurements with computational analysis. I selected a carbon steel casting as a test piece, representative of typical sand casting parts in industrial applications. The molding sand consisted of chromite sand, silica sand, iron oxide powder, resin, and hardener, with specific compositions and grain sizes. Table 1 summarizes the parameters of the sand used in my experiments.
| Category | SiO₂ | Fe₂O₃ | Cr₂O₃ | Others | Grain Size Distribution (40-70 mesh) | Grain Size Distribution (30-70 mesh) | Grain Size Distribution (30-100 mesh) |
|---|---|---|---|---|---|---|---|
| Chromite Sand | ≤1% | 20-25% | ≥46% | ≤20% | ≥75% | ≥85% | ≥95% |
| Silica Sand | ≥85% | ≤0.2% | ≤2% | ≤10% | ≥75% | ≥85% | ≥95% |
The sand mixture was prepared with urethane resin and isocyanate hardener. For chromite sand, the resin addition was 1% by weight, with hardener at 25% of resin weight. For silica sand, resin was added at 0.65% by weight, with hardener at 40% of resin weight. After mixing, the sand was compacted to achieve an average hardness of 80-90 MPa, with densities ranging from 2.5 to 2.9 kg/dm³. These conditions mimic typical production environments for sand casting parts. I then designed an experiment to measure temperature changes in the molding sand at varying distances from the casting surface. This approach allowed me to observe how heat propagates through the sand during the cooling of sand casting parts.
My experimental setup involved a large mining machinery casting made of ACM1506 low-carbon steel, a common material for sand casting parts. The casting dimensions were approximately 6400 mm × 500 mm × 700 mm, weighing about 17.5 tons, with an elliptical riser. I embedded K-type thermocouples at specific locations in the sand mold, as shown in a schematic diagram. The thermocouples were placed parallel to the heat flow direction at distances of 50 mm, 100 mm, 150 mm, and 200 mm from the casting surface. The pouring temperature was set at 1575°C, and temperatures were recorded every 10 minutes after pouring. This data collection continued until the sand temperatures stabilized, providing a comprehensive view of thermal dynamics in sand casting parts production.
The temperature data revealed significant insights. At 50 mm from the casting surface, the sand temperature increased slowly for the first 90 minutes, then rapidly rose to about 500°C between 100 and 200 minutes, before stabilizing. At 100 mm, the temperature gradually increased for 400 minutes, then spiked to around 230°C between 400 and 600 minutes. For distances of 150 mm and 200 mm, temperatures rose steadily throughout the measurement period, reaching maxima of 120°C and 80°C, respectively. These trends indicate that closer proximity to sand casting parts results in faster and more pronounced temperature changes, highlighting the low thermal conductivity of molding sand as a primary thermal resistance in sand casting. To quantify this, I analyzed the data using Fourier’s law of heat conduction, expressed as:
$$q = -k \nabla T$$
where \( q \) is the heat flux (W/m²), \( k \) is the thermal conductivity (W/m·K), and \( \nabla T \) is the temperature gradient (K/m). By applying this equation to the measured temperature profiles, I derived the thermal conductivity and specific heat of the molding sand as functions of temperature. The results are summarized in Figure 4 and Table 2 below.
| Temperature Range (°C) | Thermal Conductivity Trend | Specific Heat Trend | Notes |
|---|---|---|---|
| 20 – 400 | Decreases with temperature | Increases rapidly | Thermal conductivity minimum near 400°C |
| 400 – 600 | Increases with temperature | Increases moderately | Transition phase for sand casting parts cooling |
| 600 – 1000 | Continues to increase | Increases slowly | Stable region for high-temperature sand casting parts |
The thermal conductivity exhibited a “V”-shaped trend: it decreased as temperature rose from 20°C to 400°C, reaching a minimum at around 400°C, then increased above 400°C. This behavior can be modeled with a piecewise function. For temperatures below 400°C, the thermal conductivity \( k \) can be approximated as:
$$k(T) = k_0 – \alpha (T – T_0) \quad \text{for } T < 400^\circ \text{C}$$
where \( k_0 \) is the conductivity at a reference temperature \( T_0 \), and \( \alpha \) is a positive constant. Above 400°C, it follows:
$$k(T) = k_{400} + \beta (T – 400) \quad \text{for } T \geq 400^\circ \text{C}$$
with \( k_{400} \) being the conductivity at 400°C and \( \beta \) a positive constant. In contrast, the specific heat \( c_p \) showed a monotonic increase with temperature, but with varying rates. Below 600°C, it rose sharply, which can be described by a quadratic relation:
$$c_p(T) = c_0 + a T + b T^2 \quad \text{for } T < 600^\circ \text{C}$$
where \( c_0 \), \( a \), and \( b \) are coefficients. Above 600°C, the increase slowed, fitting a linear model:
$$c_p(T) = c_{600} + \gamma (T – 600) \quad \text{for } T \geq 600^\circ \text{C}$$
with \( c_{600} \) as the specific heat at 600°C and \( \gamma \) a small positive constant. These equations help in simulating the cooling behavior of sand casting parts more accurately. To validate these parameters, I used Magma simulation software to model the solidification process of the test casting. The sand mold was meshed with varying grid sizes: 20 mm × 20 mm × 20 mm in regions far from the casting, 5 mm × 5 mm × 5 mm near thermocouple points, and 10 mm × 10 mm × 10 mm elsewhere. This refined mesh ensured precise capture of thermal gradients around sand casting parts.
In the simulation, I initially fixed the specific heat at a constant value of 1250 J/(kg·K) and used the experimentally derived thermal conductivity. This case, labeled CASE1, yielded temperature predictions at points A1 and A2 (20 mm and 50 mm from the casting surface) that deviated from measured data. Then, in CASE2, I incorporated both variable thermal conductivity and specific heat from my study. The results showed much better agreement with experimental temperatures, as illustrated in comparative plots. This confirms that accurate thermophysical parameters are essential for reliable simulations of sand casting parts. The improvement in prediction accuracy underscores the need for site-specific parameter determination, as generic values from literature may not account for local variations in sand composition or compaction.
Further analysis involved calculating the heat transfer coefficient at the interface between the casting and sand. Using Newton’s law of cooling:
$$q = h (T_{\text{casting}} – T_{\text{sand}})$$
where \( h \) is the heat transfer coefficient (W/m²·K), I estimated \( h \) based on temperature data. This coefficient is crucial for understanding how quickly heat is dissipated from sand casting parts into the mold. My calculations indicated that \( h \) varies with temperature and sand distance, affecting the cooling rates of sand casting parts. For instance, at 50 mm, \( h \) was higher due to steeper gradients, whereas at 200 mm, it was lower, consistent with slower temperature rises. This variability must be considered in process design for sand casting parts to avoid defects like shrinkage or hot tears.
The implications of my findings are significant for the foundry industry. By using accurate thermophysical parameters, manufacturers can optimize pouring temperatures, cooling times, and mold designs for sand casting parts. This leads to improved mechanical properties, reduced scrap rates, and energy savings. For example, in producing large sand casting parts such as those for mining equipment, precise simulations can prevent over-design of risers, reducing material waste. Additionally, understanding the “V”-shaped thermal conductivity trend helps in selecting sand mixtures that enhance heat extraction where needed, such as in thick sections of sand casting parts. My study also highlights the importance of real-time monitoring during production; embedding thermocouples in molds can provide feedback for adaptive control, ensuring consistent quality across batches of sand casting parts.
To extend this research, I propose investigating the effects of different sand additives on thermophysical properties. For instance, varying the ratio of chromite to silica sand might alter conductivity and specific heat, impacting the performance for specific sand casting parts. Also, advanced measurement techniques like laser flash analysis could provide more detailed data on thermal diffusivity. Furthermore, integrating these parameters into machine learning models could enable predictive maintenance and process optimization for sand casting parts manufacturing. Such advancements would contribute to the Industry 4.0 transformation in foundries, making production of sand casting parts more efficient and sustainable.
In conclusion, my study demonstrates that the thermophysical parameters of molding sand are temperature-dependent and critical for accurate simulation of sand casting parts. The thermal conductivity follows a “V”-shaped curve with a minimum near 400°C, while specific heat increases with temperature, particularly below 600°C. These parameters, when used in numerical simulations, significantly improve prediction accuracy, offering valuable guidance for实际 production of sand casting parts. I recommend that foundries conduct similar experiments to calibrate sand properties for their specific operations, thereby enhancing the quality and reliability of sand casting parts. Future work should explore broader temperature ranges and sand compositions to build comprehensive databases for the casting industry.