The accuracy of computer simulations for the solidification and cooling of castings within a mold is fundamentally dependent on the precision of the thermophysical properties assigned to the molding sand. In the realm of professional sand casting services, where components range from simple to highly complex, reliable simulation results are indispensable for predicting solidification patterns, identifying potential shrinkage defects, and optimizing the entire casting process. However, molding sand is a porous, non-homogeneous material whose effective properties vary significantly with temperature, composition, binder type, and compaction level. Consequently, adopting generic thermophysical parameters from literature often leads to substantial inaccuracies in simulation outcomes, diminishing their value for guiding actual production. This study, conducted from a practitioner’s perspective, focuses on the experimental determination and numerical validation of the temperature-dependent thermal conductivity and specific heat of production-grade molding sand, aiming to enhance the fidelity of simulations used in industrial sand casting services.
The foundational principle governing heat transfer from the casting to the mold is Fourier’s law of heat conduction. In one-dimensional form, it is expressed as:
$$ q = -k(T) \frac{\partial T}{\partial x} $$
where \( q \) is the heat flux (W/m²), \( k(T) \) is the temperature-dependent thermal conductivity (W/(m·K)), and \( \frac{\partial T}{\partial x} \) is the temperature gradient. The transient temperature field within the sand mold is governed by the heat conduction equation:
$$ \rho c_p(T) \frac{\partial T}{\partial t} = \nabla \cdot (k(T) \nabla T) $$
where \( \rho \) is the density (kg/m³), \( c_p(T) \) is the temperature-dependent specific heat capacity (J/(kg·K)), and \( t \) is time (s). This equation highlights that both \( k(T) \) and \( c_p(T) \) are critical inputs. An inaccurate representation of these parameters directly compromises the calculated temperature evolution, which is the cornerstone of any solidification analysis in sand casting services.
Experimental Materials and Methodology
The molding sand system used in this investigation is representative of materials employed in high-quality sand casting services for large steel castings. It is a multi-layer system designed to withstand the severe thermal and mechanical loads of ferrous casting.
1. Molding Sand Composition and Preparation
The mold consisted of two distinct sand layers. A facing layer of chromite sand (15-25 mm thick) was used adjacent to the casting surface for its superior thermal stability and resistance to metal penetration. The backing material was silica sand. The chemical composition and grain size distribution are summarized in Table 1.
| Category | SiO₂ (wt%) | Fe₂O₃ (wt%) | Cr₂O₃ (wt%) | Others (wt%) | Grain Size Distribution (wt%) |
|---|---|---|---|---|---|
| Chromite Sand | ≤ 1 | 20-25 | ≥ 46 | ≤ 20 | 40-70 mesh: ≥75; 30-100 mesh: ≥95 |
| Silica Sand | ≥ 85 | ≤ 0.2 | ≤ 2 | ≤ 10 | 40-70 mesh: ≥75; 30-100 mesh: ≥95 |
The sand was bonded using a cold-box urethane resin system. For the chromite layer, resin addition was 1.0 wt% of sand, with a catalyst addition of 25 wt% relative to the resin. For the silica sand backing, resin addition was 0.65 wt% of sand, with a catalyst addition of 40 wt% relative to the resin. The mixed sand was compacted to achieve an average mold hardness between 80-90 on a scale used for production in sand casting services. The density of the compacted sand mold ranged from 2.5 to 2.9 kg/dm³.
2. Test Casting and Instrumentation
A large, low-carbon steel mining machinery casting (grade ACM1506) was selected as the test vehicle. To accurately determine the sand’s thermophysical properties, a simplified yet representative benchmark was essential. Therefore, a substantial section of the casting was modeled as a large steel block for the parameter determination study.
K-type thermocouples were embedded in the sand mold at specified distances from the casting-mold interface prior to molding. Two critical measurement locations were defined for the parameter identification process, as detailed in Table 2.
| Thermocouple Label | Distance from Casting Surface | Purpose |
|---|---|---|
| A1 | 20 mm | Primary data point for inverse calculation of sand properties. |
| A2 | 50 mm | Secondary data point for verification and refining the calculation. |
The steel was poured at a temperature of 1575°C. Temperature data from thermocouples A1 and A2 were recorded at a high sampling frequency (0.5 Hz) throughout the solidification and cooling cycle, providing the experimental temperature-time curves essential for the subsequent inverse analysis.
3. Numerical Simulation and Inverse Method
A detailed 3D model of the test casting and mold was created in a commercial casting simulation software (Magma). The computational mesh was refined in critical areas, especially near the measurement points A1 and A2 (5 mm cell size), to ensure accurate resolution of the steep temperature gradients. The initial simulations were run using standard, temperature-averaged thermophysical properties for silica sand from the software’s database.
The core of the methodology involved an iterative inverse calculation. The goal was to adjust the functions \( k(T) \) and \( c_p(T) \) for the molding sand in the simulation until the predicted temperature histories at points A1 and A2 matched the experimentally measured curves. To manage computational complexity, a sequential approach was adopted:
- Step 1 – Determine Specific Heat, \( c_p(T) \): An initial, realistic estimate for thermal conductivity \( k(T) \) was used. The specific heat function \( c_p(T) \) was then systematically varied in the simulation. The \( c_p(T) \) curve that yielded the best fit between simulated and measured temperatures at A1 and A2 was selected.
- Step 2 – Determine Thermal Conductivity, \( k(T) \): Using the optimized \( c_p(T) \) function from Step 1, the thermal conductivity function \( k(T) \) was then iteratively adjusted. The process continued until the simulated temperature curves converged with the experimental data.
This inverse method leverages the physics embedded in the numerical solver of the heat conduction equation (Equation 2) to back-calculate the material properties that must have been present to produce the observed thermal response. This data-driven approach is highly valuable for calibrating simulation inputs for specific sand casting services.
Results and Analysis of Thermophysical Properties
1. Influence of Distance on Sand Temperature
The analysis of temperatures at various distances from the casting surface reveals the insulating nature of the sand mold, a key characteristic in sand casting services that governs the cooling rate. As expected, the temperature rise is most rapid and significant closer to the interface. At a distance of 50 mm, the sand temperature rises slowly for about 90 minutes before a rapid increase to a peak near 500°C. At 100 mm, the temperature increases gradually for nearly 400 minutes before a slower rise to about 230°C. At distances of 150 mm and 200 mm, the temperature increase is very gradual over the entire measured period, reaching only about 120°C and 80°C, respectively. This pattern confirms that the primary thermal resistance in a sand mold is the molding sand itself, and the “chill effect” diminishes rapidly with distance.
2. Temperature-Dependent Thermal Conductivity, \( k(T) \)
The inversely determined thermal conductivity of the production sand mixture exhibits a distinct non-linear relationship with temperature, as plotted in Figure 4 of the source material. The trend can be described as “V-shaped”:
- From room temperature up to approximately 400°C, the effective thermal conductivity decreases with increasing temperature.
- At around 400°C, the thermal conductivity reaches a minimum value.
- For temperatures above 400°C, the thermal conductivity increases with further temperature rise.
This behavior can be attributed to competing mechanisms within the porous sand-binder system. The initial decrease may be related to the degradation of organic binders, altering the contact points between sand grains and potentially creating micro-gaps that impede heat flow. The subsequent increase above 400°C could be due to the onset of significant radiative heat transfer across pores and changes in the mineralogical structure of the sand grains or binder residues. This non-monotonic behavior is critical; using a constant average conductivity would misrepresent the heat extraction rate at different stages of cooling in sand casting services.
3. Temperature-Dependent Specific Heat, \( c_p(T) \)
The specific heat capacity of the molding sand shows a more monotonic, yet still non-linear, increase with temperature:
- From room temperature to about 600°C, the specific heat increases sharply.
- Above 600°C, the rate of increase diminishes, and the specific heat curve begins to plateau.
This increasing trend is physically reasonable. The specific heat represents the energy required to raise the temperature of a unit mass of material. As temperature increases, more energy is stored in the vibrational modes of the silicate network (sand grains) and is consumed in endothermic processes like the decomposition of binders and the removal of residual moisture. The flattening of the curve at high temperatures suggests that these processes are largely complete. For accurate simulation, accounting for this variable energy storage capacity is as important as the conductivity for predicting the timing of thermal events.
The key derived thermophysical parameters at characteristic temperatures are summarized in Table 3 below.
| Temperature (°C) | Thermal Conductivity, \( k \) (W/(m·K)) | Specific Heat, \( c_p \) (J/(kg·K)) | Notes on Trend |
|---|---|---|---|
| 25 | ~0.60 | ~900 | Baseline at room temperature. |
| 400 | ~0.35 (Min.) | ~1450 | Conductivity minimum; Specific heat rising rapidly. |
| 600 | ~0.45 | ~1600 (Plateau Start) | Inflection point for specific heat increase. |
| 800 | ~0.55 | ~1650 | Conductivity recovering; Specific heat nearly constant. |
Validation Through Numerical Simulation
The true test of the derived properties lies in their ability to improve the predictive accuracy of casting simulation software, a core tool in modern sand casting services. Two simulation cases were compared against the experimental data from thermocouples A1 and A2:
- Case 1 (Conventional Approach): Used the newly determined \( k(T) \) curve but assumed a constant average specific heat \( c_p = 1250 \, \text{J/(kg·K)} \).
- Case 2 (Proposed Approach): Used both the temperature-dependent \( k(T) \) and \( c_p(T) \) curves obtained from the inverse analysis.
The results were clear and significant. While Case 1 showed a general trend, it displayed noticeable deviations from the measured temperature history, particularly in the timing and magnitude of the temperature rise. In contrast, the temperature curves predicted by Case 2 demonstrated a remarkably close fit to the experimental data at both measurement locations. This validation confirms that the simultaneously accurate representation of both variable thermal conductivity and variable specific heat is necessary to achieve high-fidelity simulations of the thermal field in sand molds. The improvement directly translates to more reliable predictions of solidification time, thermal gradient direction, and cooling rate—all vital for defect prediction in sand casting services.
The practical implications for sand casting services are substantial. Reliable simulations enable engineers to virtually test and optimize gating and risering systems, reducing the need for costly physical trial casts. By accurately modeling how heat is stored and conducted away by the specific sand mix used, foundries can better predict shrinkage porosity locations, optimize cycle times, and improve the dimensional accuracy of castings. This level of process control and predictability is what distinguishes advanced sand casting services, allowing for the production of high-integrity components for demanding applications in mining, energy, and heavy machinery.
Conclusion
This investigation underscores the critical importance of material-specific, temperature-dependent thermophysical data for the digital modeling of casting processes. For the studied production-grade molding sand system, representative of those used in industrial sand casting services, the following key conclusions are drawn:
- The temperature evolution within the sand mold is highly sensitive to the distance from the casting interface, with thermal gradients being steepest within the first 100 mm.
- The effective thermal conductivity of the sand exhibits a pronounced “V-shaped” dependence on temperature, decreasing to a minimum around 400°C before increasing at higher temperatures. This non-linear behavior must be captured in simulations.
- The specific heat capacity of the sand increases significantly with temperature up to about 600°C, after which the rate of increase slows considerably. Using a constant value is insufficient for accurate thermal analysis.
- The inverse method combining experimental temperature measurement with iterative numerical simulation is an effective and practical approach for determining these crucial property curves.
- Implementing these validated, variable thermophysical parameters into casting simulation software dramatically improves the agreement between predicted and actual thermal histories within the mold. This enhanced accuracy directly supports the optimization of casting process design, leading to improved yield, quality, and reliability in commercial sand casting services.
Ultimately, the pursuit of accurate material data is not merely an academic exercise but a fundamental engineering practice that enhances the precision, efficiency, and competitiveness of sand casting services in a demanding manufacturing landscape.