The solidification of metals and alloys is a complex thermal event governed by the interplay between heat extraction and the release of latent heat during phase transformations. In the context of casting, the thermal history experienced at a specific location within a mould directly dictates the resultant microstructure, which in turn determines the mechanical and physical properties of the final component. For grey cast iron, a material of paramount importance in the foundry industry due to its excellent castability, machinability, and damping capacity, understanding its solidification kinetics is crucial. The analysis of cooling curves, which are time-temperature records obtained during solidification, provides a powerful and direct window into these kinetics. Through computer-aided analysis, these curves can be transformed from simple temperature logs into rich sources of quantitative data, revealing details about nucleation undercooling, growth modes, latent heat evolution, and fraction of solid formed. This methodology holds significant value for both quality control in production, enabling the prediction of microstructure and properties, and for academic research, offering critical parameters for the validation and refinement of numerical solidification simulation models.
The fundamental challenge in analyzing a cooling curve from a casting lies in deconvoluting the two simultaneous phenomena: the sensible cooling due to heat transfer to the mould, and the thermal effects caused by the release of latent heat. Even within a single casting of grey cast iron, different sections cool at different rates, producing unique local thermal histories. This work delves into the mathematical foundation and practical application of computer-aided cooling curve analysis (CA-CCA), focusing specifically on the solidification of a hypoeutectic grey cast iron casting. By applying this technique to a real-world casting, we aim not only to demonstrate a robust analytical method but also to deepen the mechanistic understanding of the microstructural transformations in grey cast iron.
Mathematical Foundation of Cooling Curve Analysis
The analysis begins with a macroscopic energy balance for the solidifying metal. For a small casting or a small-volume element within a sand mould where temperature gradients can be considered negligible (i.e., assuming a lumped capacitance system), the heat balance can be expressed as:
$$
\frac{dQ_s}{dt} = V \rho C_p \frac{dT}{dt} + h’A(T – T_0)
$$
where \( Q_s \) is the latent heat released by the solidifying metal (J), \( V \) is the volume of the cooling body (m³), \( \rho \) is the density (kg/m³), \( C_p \) is the specific heat capacity (J/kg·K), \( T \) is the temperature (K), \( t \) is time (s), \( h’ \) is an effective heat transfer coefficient (W/m²·K), \( A \) is the surface area for heat exchange (m²), and \( T_0 \) is the ambient/environment temperature (K).
This equation states that the rate of latent heat generation (left side) equals the sum of the rate of change in sensible heat of the metal and the rate of heat loss to the surroundings. Rearranging this equation gives the general cooling rate during solidification:
$$
\frac{dT}{dt} = \frac{1}{V \rho C_p} \left[ \frac{dQ_s}{dt} – h’A(T – T_0) \right]
$$
The core idea of CA-CCA is to compare the actual cooling curve, which includes the effect of latent heat, with a hypothetical “baseline” curve that would have been followed if no phase transformation occurred. This baseline is called the “zero curve” or “Newtonian cooling curve.” If no phase change occurs (\(dQ_s/dt = 0\)), Equation (2) simplifies to the Newtonian cooling law:
$$
\left( \frac{dT}{dt} \right)_{nc} = – \frac{h’A}{V \rho C_p} (T – T_0)
$$
where the subscript \( nc \) denotes the “no-phase-change” condition. The actual cooling rate measured from the casting is denoted with the subscript \( c \). The key assumption in the method is that the release of latent heat does not significantly alter the external heat transfer conditions between the metal and the mould at any given moment. This is typically ensured by using a small test sample (like a thermal analysis cup) or analyzing a small section of a casting, making the lumped capacitance assumption more valid.
By subtracting the Newtonian cooling rate (Equation 3) from the actual cooling rate (derived from Equation 2), we can isolate the thermal effect due solely to latent heat release:
$$
\frac{dQ_s}{dt} = V \rho C_p \left[ \left( \frac{dT}{dt} \right)_{nc} – \left( \frac{dT}{dt} \right)_{c} \right]
$$
Integrating this rate over the total solidification time \( t_s \) yields the total latent heat released, \( Q_s \):
$$
Q_s = \int_{0}^{t_s} \frac{dQ_s}{dt} dt = V \rho \int_{0}^{t_s} \left[ \left( \frac{dT}{dt} \right)_{nc} – \left( \frac{dT}{dt} \right)_{c} \right] C_p \, dt
$$
If we consider the specific latent heat \( L \) (J/kg), the total released heat is \( Q_s = m L = V \rho L \). Therefore, the specific latent heat can be calculated from the cooling curves:
$$
L = \frac{1}{V \rho} \int_{0}^{t_s} \frac{dQ_s}{dt} dt = \int_{0}^{t_s} \left[ \left( \frac{dT}{dt} \right)_{nc} – \left( \frac{dT}{dt} \right)_{c} \right] C_p \, dt
$$
Equation (6) shows that the total latent heat per unit mass is proportional to the integrated area between the baseline (Newtonian) cooling rate curve and the actual cooling rate curve, weighted by the specific heat capacity. Assuming a direct proportionality between the amount of solid formed and the latent heat released, the solid fraction \( F_s \) at any time \( t \) can be derived:
$$
F_s(t) = \frac{1}{L} \int_{0}^{t} \left[ \left( \frac{dT}{dt} \right)_{nc} – \left( \frac{dT}{dt} \right)_{c} \right] C_p \, dt
$$
Differentiating this expression gives the solid formation rate, a critical parameter for microstructure modeling:
$$
\frac{dF_s}{dt} = \frac{C_p}{L} \left[ \left( \frac{dT}{dt} \right)_{nc} – \left( \frac{dT}{dt} \right)_{c} \right]
$$
In practice, the actual cooling curve \( T_c(t) \) is obtained experimentally and can be easily digitized and differentiated numerically. The challenge is constructing the Newtonian baseline \( (dT/dt)_{nc} \). An approximate expression can be obtained by integrating and then differentiating Equation (3). For a body cooling under Newtonian conditions from an initial temperature \( T_{pour} \), the temperature evolution is:
$$
T_{nc}(t) = T_0 + (T_{pour} – T_0) \exp\left(-\frac{h’A}{V \rho C_p} t\right)
$$
Differentiating this gives the baseline cooling rate. Often, the modulus \( M = V/A \) is used for characterization. The cooling constant \( K = h’/(\rho C_p M) \) can be estimated from the initial slope of the actual cooling curve before any transformation begins. Therefore, a practical form of the baseline cooling rate is:
$$
\left( \frac{dT}{dt} \right)_{nc} \approx -K (T_{pour} – T_0) \exp(-K t)
$$
For accurate calculations, the temperature dependence of the specific heat capacity \( C_p \) for grey cast iron must be considered. A typical linear relationship found in the literature is:
$$
C_p(T) = 727.98 – 0.11047 \, T \quad \text{(J/kg·K)}
$$
where \( T \) is in degrees Celsius. This equation is integrated into the calculations for both latent heat and solid fraction.
Experimental Methodology and Data Acquisition
To demonstrate the CA-CCA technique, a hypoeutectic grey cast iron casting with a nominal composition equivalent to HT300 (Carbon Equivalent, CE ≈ 3.5%) was investigated. The melt was prepared in a cupola furnace, subjected to inoculation treatment, and poured into a dry sand mould. The casting of interest was a chuck body, with six cavities per mould box.
Temperature-time data was acquired directly from the casting during solidification. A thermocouple (typically type K) was embedded at a critical location—the center of the chuck body—to capture the thermal history representative of a slowly cooling section. The analog signal from the thermocouple was recorded using both a multi-channel data acquisition system and a high-speed function recorder to ensure accuracy and redundancy. The recorded cooling curve was then digitized at fine intervals to facilitate numerical analysis.
The core of the analysis was performed using a dedicated program written in Fortran. The algorithm executed the following key steps:
- Data Smoothing: Applied to the raw digitized temperature data to reduce high-frequency noise that could be amplified during differentiation.
- Numerical Differentiation: Calculated the first derivative \( (dT/dt)_c \) and the second derivative \( (d^2T/dt^2)_c \) of the cooling curve using finite difference methods.
- Feature Identification: Used the second derivative curve to precisely identify critical points on the cooling curve, such as the onset of primary austenite formation and the eutectic recalescence.
- Baseline Construction: Generated the Newtonian baseline cooling rate \( (dT/dt)_{nc} \) based on Equation (10), using the modulus of the test section and an estimated cooling constant.
- Integration and Calculation: Computed the area difference between the baseline and actual cooling rate curves, applying the temperature-dependent \( C_p \), to yield latent heat evolution \( Q_s(t) \) and solid fraction \( F_s(t) \) via Equations (7) and (8).
Analysis of Results for Grey Cast Iron Solidification
The primary output of the numerical differentiation is presented conceptually in the figure below. The first derivative curve \( (dT/dt)_c \) clearly shows the influence of latent heat. The second derivative curve \( (d^2T/dt^2)_c \) is particularly useful for pinpointing the exact moments of phase change initiation.
Characteristic Points:
- Point C₁: The first significant peak on the left side of the second derivative curve, occurring at approximately 1055°C. This corresponds to the nucleation and onset of growth of primary austenite dendrites from the liquid. The associated temperature is around 1240°C.
- Point C₂: A major inflection identified at a later time (e.g., 223s). This marks the beginning of the eutectic transformation. At this point, the first derivative \( (dT/dt)_c \) begins to increase (becomes less negative) because the release of eutectic latent heat starts to counteract the Newtonian cooling. Shortly after C₂, a significant release of latent heat causes a temperature rise known as recalescence, creating a distinct plateau or hump on the cooling curve itself.
The calculated solid fraction \( F_s \) as a function of time is shown in the following conceptual plot. The curve exhibits a characteristic sigmoidal shape, starting slowly, accelerating during the main eutectic reaction, and tapering off at the end of solidification.
An important finding from the derivative of the solid fraction curve \( dF_s/dt \) is its near-linear behavior during distinct phases. For the analyzed hypoeutectic grey cast iron, the solid formation rate can be approximated by a linear function of time in two segments:
$$
\frac{dF_s}{dt} = a + b t
$$
For the primary austenite precipitation segment, regression analysis yielded parameters \( a_1 = -0.002173 \, s^{-1} \) and \( b_1 = 0.000061 \, s^{-2} \). For the subsequent eutectic segment, the parameters were \( a_2 = 0.00054 \, s^{-1} \) and \( b_2 = -0.0000061 \, s^{-2} \). This linear approximation, valid with a high confidence level (α = 95%), provides a simple yet valuable empirical model for feeding into macroscopic solidification simulation software for grey cast iron.
A critical visual output is the overlay of the cooling curve, its first derivative, and the constructed Newtonian baseline derivative. The area between the baseline and the actual first derivative curve is directly proportional to the latent heat released. This area can be segmented:
- The area from C₁ to the minimum of the first derivative curve (the “valley”) corresponds to the latent heat released during primary austenite formation (\(Q_{γ}\)) and defines the primary austenite fraction.
- The area from this valley to the end of solidification corresponds to the latent heat released during the eutectic transformation (\(Q_{E}\)) and defines the eutectic fraction.
The minimum (valley) in the first derivative curve effectively marks the transition where the rate of eutectic heat release begins to dominate over the primary phase release, often interpreted as the end of significant primary growth.
The quantitative results from the CA-CCA of the chuck casting are summarized and compared with literature data for similar grey cast iron in the table below. The comparison highlights the influence of composition and cooling conditions.
| Carbon Equivalent (CE %) | Primary Austenite Latent Heat, Lγ (kJ/kg) | Eutectic Latent Heat, LE (kJ/kg) | Total Latent Heat, L (kJ/kg) | Austenite Fraction, Fγ | Eutectic Fraction, FE |
|---|---|---|---|---|---|
| 3.4 (Reference Data*) | 92.5 | 172.2 | 264.7 | 0.35 | 0.65 |
| 3.5 (This Work) | 75.5 | 209.4 | 285.1 | 0.26 | 0.74 |
*Reference data typically obtained from standardized thermal analysis cups with faster cooling.
The differences are instructive. The higher total latent heat in the present work likely stems from the slower cooling in a sand mould compared to the metallic cup used for reference data. Slower cooling allows for more complete diffusion and potentially a higher degree of graphitization, which releases more latent heat than the formation of cementite. The lower primary austenite fraction and correspondingly higher eutectic fraction for the higher CE (3.5%) are consistent with the phase diagram, as a higher carbon equivalent moves the composition closer to the eutectic point.
Discussion and Industrial Implications
The computer-aided cooling curve analysis method presented here transcends simple temperature recording. It is a diagnostic tool that quantifies the dynamics of solidification. For grey cast iron, this is particularly valuable because its properties are intimately linked to the graphite morphology and matrix structure, both of which are controlled by solidification parameters.
The ability to determine real-time solid fraction and its rate of change is a direct input for feeding simulation algorithms and for predicting shrinkage porosity susceptibility. The distinct signatures of the primary and eutectic reactions allow for the assessment of inoculation effectiveness; a well-inoculated iron will show a smaller undercooling before the eutectic recalescence and a more pronounced temperature plateau. Furthermore, by analyzing curves from different locations in a casting, one can map the variation in local solidification time and cooling rate, which correlates with hardness and microstructure gradients.
In an industrial setting, this technique can be integrated into a quality monitoring system. For high-volume production of critical grey cast iron components like brake discs, engine blocks, or hydraulic parts, real-time thermal analysis from a test coupon poured from each ladle can provide immediate feedback on the melt quality (CE, inoculation level) before the main moulds are poured. This proactive approach can significantly reduce scrap rates.
Conclusion
This detailed exploration has demonstrated that Computer-Aided Cooling Curve Analysis (CA-CCA) is a powerful and practical methodology for investigating the solidification of grey cast iron. By establishing a sound mathematical framework based on heat balance and employing numerical differentiation and integration techniques, the latent heat evolution and solid fraction development during solidification can be accurately derived from a simple temperature-time recording.
The analysis of a hypoeutectic grey cast iron chuck casting revealed characteristic cooling curve features corresponding to primary austenite and eutectic graphite-austenite formation. The solid formation rate was found to be approximately linear within each major transformation stage, providing a useful simplified model. The quantitative results, including the partitioning of latent heat between primary and eutectic phases, offer valuable data for calibrating and validating more complex numerical simulation models of grey cast iron solidification.
Ultimately, CA-CCA serves as a critical bridge between empirical foundry practice and scientific analysis. It transforms the cooling curve from a passive log into an active source of information, enabling better control, prediction, and optimization of the production processes for high-quality grey cast iron castings.