Gray cast iron remains one of the most widely used casting materials globally, accounting for a significant portion of cast metal production. Its favorable properties, including excellent castability, good machinability, damping capacity, and low cost, make it indispensable for automotive components and industrial machinery. The final microstructure and, consequently, the mechanical properties of a gray cast iron casting are decisively influenced by its solidification process. Variables such as chemical composition, pouring temperature, and the local cooling rate within the mold alter the nucleation and growth kinetics of the constituent phases—primary austenite, graphite, and the eutectic mixture. While the effect of composition has been studied extensively, a systematic investigation linking specific solidification conditions, revealed through thermal analysis, to the resultant microstructure and mechanical properties for a fixed alloy composition is less common. This study aims to bridge that gap by employing differential thermal analysis to investigate the solidification characteristics of a hypoeutectic gray cast iron under systematically varied cooling conditions. We correlate key thermal analysis parameters with quantitatively assessed microstructural features and mechanical strength, establishing predictive models that can enhance quality control in foundry practice.
1. Introduction and Background
The foundry industry continuously seeks methods to predict and control the quality of cast components efficiently. Thermal analysis has emerged as a powerful, rapid technique for on-line assessment of molten metal quality. The principle is straightforward: as a metal solidifies, phase transformations release latent heat, causing characteristic inflections or plateaus in the cooling curve. By analyzing this temperature-time curve and its derivative, critical temperatures associated with phase nucleation and growth can be identified. For gray cast iron, this allows for the estimation of carbon equivalent, prediction of eutectic cell count, and assessment of graphite morphology and mechanical properties.
Despite its established use, the application of thermal analysis often relies on empirical correlations that may not fully account for the interplay between process parameters like pouring temperature and cooling rate. For a given melt composition, these parameters govern the undercooling, recalescence, and solidification time, all of which shape the microstructure. A finer, more uniform distribution of graphite flakes, a refined austenitic dendritic network, and a higher number of smaller eutectic cells are generally associated with improved tensile strength. However, quantifying how specific thermal analysis features evolve with changing cooling conditions and linking them directly to these microstructural metrics and ultimate strength limits requires a focused experimental approach.
This work focuses on a hypoeutectic gray cast iron with a fixed chemical composition. Two series of experiments were conducted: one varying the pouring temperature using standard sampling cups, and another varying the cooling rate using a step-shaped casting with thermocouples placed in sections of different thicknesses. We performed differential thermal analysis to extract characteristic parameters, conducted metallographic examination to quantify graphite morphology, austenite dendrite arm spacing, and eutectic cell size, and finally performed tensile and compression tests. The core objectives are to: (1) analyze the influence of pouring temperature and cooling rate on thermal analysis signatures, (2) validate microstructural predictions based on these signatures, and (3) establish nonlinear relationships between key thermal analysis parameters and the mechanical strength limits of the gray cast iron.
2. Experimental Methodology
2.1 Material and Melt Preparation
The experiments utilized an un-inoculated hypoeutectic gray cast iron. The charge consisted of foundry pig iron, high-quality steel scrap, and returns. Melting was carried out in a 3-ton medium-frequency induction furnace with an acid lining. The melt was superheated to 1510 ± 10°C and held for 3-5 minutes before tapping at a temperature above 1480°C. The chemical composition of the final iron is given in Table 1.
| Element | C | Si | Mn | P | S | CE |
|---|---|---|---|---|---|---|
| Content (wt.%) | 3.12 | 1.73 | 0.83 | 0.023 | 0.018 | 3.63 |
Note: Carbon Equivalent, CE = C + 0.3(Si+P) – 0.4(Mn+S).
2.2 Thermal Analysis Experiments
Two distinct experimental setups were employed to isolate the effects of pouring temperature and cooling rate.
Pouring Temperature Series: Standard square sampling cups (35 mm x 35 mm x 40 mm) with an anti-graphitizing coating were used. K-type thermocouples (0.6 mm diameter) were fixed horizontally at the geometric center of each cup. Iron was poured at five different temperatures, approximately 20 seconds apart, to achieve a range from 1217°C to 1241°C. One cup was poured only halfway at 1230°C to serve as a high cooling rate reference.
Cooling Rate Series: A step-shaped casting (green sand mold) with six sections of varying thickness was designed to produce a range of cooling rates. The dimensions and thermocouple positions are detailed in Table 2. K-type thermocouples were embedded in the sand mold at the center of each section. All thermocouples were connected via compensation cables to a NI CompactDAQ data acquisition system, recording temperature data at a high frequency until the castings cooled to approximately 1000°C.
| Step No. | Dimensions (L×W×H, mm) | Thermocouple Position (mm) |
|---|---|---|
| S1 | 180 × 25 × 5 | (90, 12.5, 2.5) |
| S2 | 180 × 35 × 10 | (90, 17.5, 5) |
| S3 | 180 × 40 × 15 | (90, 20, 7.5) |
| S4 | 180 × 50 × 20 | (90, 25, 10) |
| S5 | 180 × 55 × 30 | (90, 27.5, 15) |
| S6 | 180 × 60 × 45 | (90, 30, 22.5) |
2.3 Thermal Analysis Parameter Definition
The recorded cooling curves (T-t) were smoothed and differentiated numerically to obtain the first derivative (dT/dt). Key characteristic points were identified on both curves, as illustrated in Figure 1 and defined in Table 3. Beyond traditional parameters, we introduced two additional parameters: the eutectic solidification time ($\Delta t_{ES}$) and the difference between the equilibrium eutectic temperature and the final solidification temperature ($\Delta T_2$), to better describe the solidification process.
| Symbol | Definition |
|---|---|
| $T_{max}$ | Maximum temperature recorded (°C) |
| $T_{AL}$ | Temperature at start of primary austenite formation (°C) |
| $T_{SEF}$ | Temperature at start of eutectic formation (nucleation) (°C) |
| $T_{EU}$ | Minimum temperature during eutectic reaction (°C) |
| $T_{ER}$ | Maximum recalescence temperature after eutectic undercooling (°C) |
| $T_{EF}$ | Temperature at end of eutectic solidification (°C) |
| $T_E$, $T_E’$ | Stable and metastable eutectic equilibrium temperatures (°C) |
| $\Delta T_R$ | Eutectic recalescence: $T_{ER} – T_{EU}$ (°C) |
| $\Delta T_1$ | Eutectic nucleation undercooling: $T_E – T_{EU}$ (°C) |
| $\Delta T_2$ | Actual solidification end temperature difference: $T_E – T_{EF}$ (°C) |
| $\Delta T_{AE}$ | $T_{AL} – T_{SEF}$ (°C) |
| $\Delta t_{ES}$ | Eutectic solidification time: $t(T_{SEF}) – t(T_{EF})$ (s) |
| $V_1$ | Cooling rate from $T_{AL}$ to $T_{SEF}$ (°C/s) |
| $V_2$ | Overall cooling rate from $T_{AL}$ to $T_{EF}$ (°C/s) |
2.4 Microstructural and Mechanical Characterization
Samples for microstructural analysis were sectioned from the thermal analysis cups and from the center of each step in the staircase casting. Standard metallographic preparation involved grinding, polishing, and etching. Graphite morphology was observed on polished samples. To reveal the austenitic dendritic structure and eutectic cells, samples were etched with an ammonium persulfate solution and a CuSO₄-HCl solution, respectively. Quantitative analysis was performed using the linear intercept method (according to GB/T6394-2002) to determine the average eutectic cell diameter, primary dendrite arm spacing (DAS), and secondary dendrite arm spacing (SDAS).
For mechanical testing, standard tensile and compression specimens were machined from the material adjacent to the thermocouple locations in the respective castings, following GB/T9439-2010. Tensile tests were performed at a crosshead speed of 1 mm/min, and compression tests at 0.50 mm/min, using a Shimadzu AG-IC electronic universal testing machine. The ultimate tensile strength (UTS) and ultimate compressive strength (UCS) were recorded.
3. Results: Influence of Cooling Conditions on Solidification Signatures
3.1 Effect of Pouring Temperature
The cooling curves and their derivatives for the cup samples poured at different temperatures showed systematic shifts. As the pouring temperature decreased, the total solidification time shortened, and the eutectic reaction commenced earlier. The sample poured only halfway (at 1230°C) exhibited a distinctly lower recalescence temperature and a much shorter solidification time due to its significantly higher cooling rate, highlighting the dominant effect of cooling rate over the small changes in pouring temperature.
Key thermal analysis parameters extracted from these curves are summarized in Table 4. The start of austenite formation ($T_{AL}$) and the start of eutectic formation ($T_{SEF}$) showed a slight increasing trend with higher pouring temperature. The parameter $\Delta T_{AE}$, representing the temperature interval for primary austenite growth, increased notably only at the highest pouring temperature (1241°C), suggesting a complex interaction. The cooling rate during the primary phase formation ($V_1$) varied between 1.08 and 1.52°C/s, with no clear correlation to pouring temperature.
| Pour Temp. (°C) | $T_{AL}$ (°C) | $T_{SEF}$ (°C) | $T_{EU}$ (°C) | $T_{ER}$ (°C) | $\Delta T_R$ (°C) | $\Delta T_1$ (°C) | $\Delta T_2$ (°C) |
|---|---|---|---|---|---|---|---|
| 1241 | 1212 | 1190 | 1147 | 1153 | 6 | ~15 | ~61 |
| 1234 | 1209 | 1187 | 1146 | 1151 | 5 | ~14 | ~68 |
| 1230* | 1200 | 1178 | 1140 | 1145 | 5 | ~12 | ~76 |
| 1225 | 1206 | 1186 | 1145 | 1148 | 3 | ~13 | ~62 |
| 1217 | 1200 | 1170 | 1140 | 1143 | 3 | ~12 | ~87 |
* Cup poured halfway, higher cooling rate.
The eutectic undercooling parameters $T_{EU}$ and $T_{ER}$ were relatively insensitive to the small changes in pouring temperature but were significantly lower for the high-cooling-rate sample. The recalescence $\Delta T_R$ increased from 3°C to 6°C with increasing pouring temperature. The eutectic nucleation undercooling $\Delta T_1$ (estimated relative to the stable eutectic temperature) showed a tendency to increase with higher pouring temperature. More importantly, the actual solidification end temperature difference $\Delta T_2$ decreased substantially as pouring temperature increased, indicating a more uniform final stage of solidification and potentially fewer micro-shrinkage defects. The eutectic solidification time $\Delta t_{ES}$ also increased with higher pouring temperature, allowing more time for solute diffusion.
3.2 Effect of Cooling Rate (Step Casting)
The effect of cooling rate, isolated as much as possible via the step casting geometry, was profound. The cooling curve for the thinnest section (S1) showed no clear austenite plateau and a very brief eutectic reaction, while the thickest section (S6) displayed a pronounced primary austenite plateau and a prolonged eutectic solidification period. The extracted parameters are listed in Table 5. It is important to note that $T_{max}$ varied between steps due to heat loss during filling; thus, the analysis focuses on the trend with decreasing cooling rate (increasing section thickness).
| Step (Cooling Rate) | $T_{AL}$ (°C) | $T_{SEF}$ (°C) | $\Delta T_{AE}$ (°C) | $\Delta T_R$ (°C) | $\Delta T_1$ (°C) | $\Delta T_2$ (°C) | $\Delta t_{ES}$ (s) |
|---|---|---|---|---|---|---|---|
| S1 (High) | – | – | – | 0 | ~52 | ~82 | 17 |
| S2 | 1210 | 1193 | 17 | 1 | ~13 | ~32 | 58 |
| S3 | 1237 | 1229 | 8 | 1 | ~3 | ~16 | 189 |
| S4 | 1246 | 1229 | 17 | 3 | ~ -1 | ~14 | 343 |
| S5 | 1245 | 1231 | 14 | 6 | ~ -1 | ~10 | 583 |
| S6 (Low) | 1248 | 1232 | 16 | 8 | ~ -3 | ~6 | 720 |
$T_{AL}$ and $T_{SEF}$ increased significantly as the cooling rate decreased (moving from S2 to S6). $\Delta T_{AE}$ fluctuated between 14°C and 17°C without a clear trend, suggesting it is less sensitive to cooling rate alone. The most striking changes were observed in $\Delta T_R$ and $\Delta T_1$. The recalescence $\Delta T_R$ increased monotonically from 1°C to 8°C as cooling rate decreased. The eutectic nucleation undercooling $\Delta T_1$ decreased dramatically, even becoming negative for the slower-cooling sections (S4-S6). A negative $\Delta T_1$ indicates that the undercooling $T_E – T_{EU}$ is less than zero, meaning $T_{EU}$ is above $T_E$. This typically occurs when the metastable (carbidic) eutectic temperature $T_E’$ is the relevant equilibrium temperature, suggesting a tendency for cementite formation at the onset of eutectic solidification in very slow-cooling conditions for this specific alloy. Conversely, the high undercooling in section S1 strongly favors graphite nucleation. The parameter $\Delta T_2$ decreased markedly with slower cooling, and the eutectic solidification time $\Delta t_{ES}$ increased exponentially.
4. Results: Correlation with Microstructure
4.1 Graphite Morphology
The microstructure analysis confirmed the predictions from thermal analysis. In the pouring temperature series, higher pouring temperatures resulted in coarser, straighter, type A graphite flakes. Lower pouring temperatures led to finer, more curved flakes, with an increased presence of undercooled (type D/E) graphite, especially in the sample with the highest cooling rate (half-filled cup).
In the cooling rate series, the transformation was even more pronounced. The slowly cooled section S6 exhibited predominantly coarse, well-formed type A graphite. As the cooling rate increased (S4, S3), the graphite became finer, more branched, and types D and E became increasingly common. In the fastest-cooled section S1, the structure was dominated by fine, interdendritic undercooled graphite (types D/E). This progression is directly linked to the increasing $\Delta T_R$ and $\Delta T_1$ values with slower cooling, which promote the growth of larger, more stable graphite flakes.
4.2 Austenite Dendritic Structure
The morphology of the primary austenite dendrites was severely affected by the cooling rate. In the thick section S6, the dendrites were coarse with large primary and secondary arm spacings, and the network was less developed. As cooling rate increased, the dendritic network became much finer and more complex. Quantitative measurements confirmed this:
- Primary Dendrite Arm Spacing (DAS) decreased from ~150.6 μm (S6) to ~73.0 μm (S1).
- Secondary Dendrite Arm Spacing (SDAS) decreased from ~80.3 μm (S6) to ~24.3 μm (S1).
The relationship between secondary dendrite arm spacing and local solidification time or cooling rate often follows a power law, such as $SDAS = k \cdot (t_f)^n$ or $SDAS = k \cdot (R)^{-n}$, where $R$ is the cooling rate. Our data aligns with this fundamental solidification principle. Pouring temperature had a less dramatic but noticeable effect, with lower temperatures promoting a slightly finer dendritic structure due to a higher effective undercooling at the onset of solidification.
4.3 Eutectic Cells
The size and number of eutectic cells were inversely related to the cooling rate. The average eutectic cell diameter increased from approximately 192 μm in the fastest-cooled section (S1) to 489 μm in the slowest-cooled section (S6). Higher cooling rates (and the associated larger $\Delta T_1$) provide a greater driving force for nucleation, resulting in a higher number of eutectic cells that grow to a smaller final size before impingement. In the pouring temperature series, a similar trend was observed where higher pouring temperatures (associated with slightly lower effective cooling rates and smaller $\Delta T_1$) led to larger eutectic cells (e.g., ~318 μm at 1241°C vs. ~254 μm at 1217°C).
5. Results: Correlation with Mechanical Properties
5.1 Ultimate Tensile Strength (UTS)
Tensile tests were performed on specimens from the step casting. The UTS increased consistently with increasing cooling rate (decreasing section thickness), as shown in Table 6. This is a direct consequence of the microstructural refinement: finer graphite, smaller eutectic cells, and a finer austenitic dendritic network all contribute to higher strength by reducing stress concentration sites and impeding crack propagation.
| Specimen Location (Step) | UTS (MPa) | $\Delta T_R$ (°C) |
|---|---|---|
| S6 (Center) | 210.2 | 8 |
| S5 | 233.6 | 6 |
| S4 | 243.5 | 3 |
| S3 | 245.2 | 1 |
A strong correlation was found between the UTS and the thermal analysis parameter $\Delta T_R$. As $\Delta T_R$ decreased (indicating less recalescence, associated with finer graphite formed under higher undercooling), the UTS increased. A nonlinear regression model was established:
$$ \sigma_{UTS} = 246.10 – \frac{0.53124}{\exp(\frac{\Delta T_R}{1.89816})} $$
where $\sigma_{UTS}$ is the ultimate tensile strength in MPa. This model allows for an estimation of tensile strength directly from the thermal analysis cooling curve of a gray cast iron sample solidified under similar conditions.
5.2 Ultimate Compressive Strength (UCS)
Compression tests were conducted on specimens from the cup samples with different pouring temperatures. The UCS decreased with increasing pouring temperature (Table 7), which correlates with the coarsening of the microstructure. While graphite flakes severely weaken the material in tension, they have a lesser effect under compression, where the matrix strength plays a more dominant role. However, a coarser microstructure generally implies lower overall strength.
| Pouring Temperature (°C) | UCS (MPa) | $\Delta T_1$ (°C)* |
|---|---|---|
| 1217 | 724.7 | ~12 |
| 1225 | 719.3 | ~13 |
| 1230 (Half) | 768.0 | ~12 |
| 1234 | 705.5 | ~14 |
| 1241 | 686.3 | ~15 |
* Estimated values relative to stable eutectic temperature.
A correlation was observed between the UCS and the eutectic nucleation undercooling parameter $\Delta T_1$. Higher $\Delta T_1$ (associated with higher pouring temperature in this series) led to lower compressive strength. The following nonlinear relationship was derived:
$$ \sigma_{UCS} = 726.59 – \frac{40.308}{1 + \exp\left(\frac{\Delta T_1 – 17.072}{0.6307}\right)} $$
where $\sigma_{UCS}$ is the ultimate compressive strength in MPa. This sigmoidal model suggests a gradual transition in the sensitivity of compressive strength to changes in eutectic undercooling.
6. Discussion
This study successfully demonstrates the integrated use of thermal analysis, metallography, and mechanical testing to deconvolute the effects of cooling conditions on hypoeutectic gray cast iron. The thermal analysis parameters $\Delta T_1$ and $\Delta T_R$ proved to be particularly sensitive and informative. $\Delta T_1$, the eutectic nucleation undercooling, is a direct measure of the driving force for graphite nucleation. High $\Delta T_1$ values promote a larger number of nuclei, leading to a finer eutectic structure, which is beneficial for strength. Our results show that high cooling rates generate large $\Delta T_1$, while low cooling rates can even result in negative $\Delta T_1$, hinting at a shift towards the metastable system. $\Delta T_R$, the recalescence magnitude, is indicative of the growth process following nucleation. A large $\Delta T_R$ is typically associated with the growth of coarse, stable graphite flakes in a near-equilibrium manner, which occurs under slow cooling. A small $\Delta T_R$ indicates rapid growth in a highly undercooled melt, leading to fine, undercooled graphite.
The introduced parameter $\Delta T_2$ effectively captured the uniformity of the final stages of solidification. A smaller $\Delta T_2$ signifies that the last liquid to solidify does so at a temperature closer to the eutectic temperature, implying less micro-segregation and a lower propensity for shrinkage porosity. This parameter decreased with both lower cooling rates and higher pouring temperatures, conditions that allow for better solute diffusion.
The established nonlinear models for UTS and UCS provide a practical tool for foundries. By performing a simple thermal analysis test with a standardized sample cup, key parameters like $\Delta T_R$ and $\Delta T_1$ can be extracted. These can then be used to predict the expected tensile and compressive strength ranges for the subsequent castings, assuming other conditions (melting practice, molding material) are consistent. This moves beyond traditional carbon equivalent estimation towards a more direct prediction of performance.
A limitation of the step-casting experiment is the unavoidable variation in $T_{max}$ between sections, which superimposes the effect of initial heat content on the effect of cooling rate. Future work could employ controlled cooling furnaces or different mold materials to achieve different cooling rates from an identical pouring temperature. Furthermore, the models presented are specific to the alloy composition studied. Expanding this methodology to create a database for different grades of gray cast iron would be a valuable endeavor.
7. Conclusion
1. The solidification of hypoeutectic gray cast iron is profoundly influenced by cooling conditions, which are clearly reflected in thermal analysis cooling curves and their derivatives. Decreasing the cooling rate increases characteristic temperatures ($T_{AL}$, $T_{SEF}$, $T_{ER}$), increases recalescence $\Delta T_R$, decreases eutectic nucleation undercooling $\Delta T_1$, and significantly prolongs eutectic solidification time $\Delta t_{ES}$.
2. The microstructural evolution correlates directly with the thermal analysis parameters. Higher cooling rates (high $\Delta T_1$, low $\Delta T_R$) produce finer, more curved graphite (increased D/E types), a finer and more complex austenitic dendritic network (reduced DAS and SDAS), and a larger number of smaller eutectic cells. Increasing pouring temperature has a similar but less pronounced coarsening effect.
3. The mechanical properties are a direct consequence of this microstructural refinement. Ultimate tensile strength increases with cooling rate. A predictive nonlinear model was established between UTS and the recalescence parameter: $\sigma_{UTS} = 246.10 – 0.53124 / \exp(\Delta T_R / 1.89816)$.
4. Ultimate compressive strength decreases with increasing pouring temperature, correlating with increased $\Delta T_1$. A predictive sigmoidal model was established: $\sigma_{UCS} = 726.59 – 40.308 / [1 + \exp((\Delta T_1 – 17.072)/0.6307)]$.
This research confirms that thermal analysis is not merely a tool for compositional check but a powerful technique for predicting the as-cast microstructure and mechanical performance of gray cast iron. By understanding the links between process parameters, thermal analysis signatures, and final properties, foundries can better optimize their processes to achieve consistent, high-quality castings.