In the field of advanced manufacturing, steel castings play a pivotal role due to their design flexibility, ability to form complex shapes, and high material utilization. Particularly in sectors such as shipbuilding, automotive, engineering machinery, and aerospace, large-scale steel castings are indispensable. However, the inherent structural and wall-thickness variations in these steel castings often lead to severe segregation and microstructural inhomogeneity, which can detrimentally affect mechanical properties, especially low-temperature toughness. Therefore, optimizing heat treatment processes is crucial to enhance the performance of steel castings. This study focuses on a 10CrNiCu low-alloy steel casting, employing an orthogonal experimental design to systematically investigate the effects of normalizing cycles, tempering temperature, and tempering time on mechanical properties and microstructure. The goal is to develop an optimized heat treatment protocol that balances strength and toughness, ensuring the reliability of steel castings in demanding applications.
The significance of steel castings cannot be overstated; they enable the production of components with intricate geometries that are difficult or impossible to achieve through other methods. However, the casting process often introduces defects like microsegregation, which can lead to localized variations in alloying elements and, consequently, non-uniform microstructures. This is particularly problematic for low-alloy steels, where elements such as Cr, Ni, and Cu are added to improve hardenability and corrosion resistance. In steel castings, these elements tend to segregate, forming bands rich in alloying content that transform into undesirable phases like bainite upon heat treatment, compromising toughness. Thus, post-casting heat treatments, including normalizing and tempering, are essential to refine the microstructure and homogenize the properties of steel castings.
To address these challenges, I designed an orthogonal experiment, which is a statistical method that allows for the efficient evaluation of multiple factors and levels with minimal experimental runs. This approach is ideal for optimizing complex processes like heat treatment for steel castings, where interactions between parameters can be significant. The factors selected were the number of normalizing cycles (1, 2, or 3), tempering temperature (570°C, 600°C, or 630°C), and tempering time (2 h, 4 h, or 8 h). These factors were chosen based on prior knowledge that normalizing refines grains, while tempering relieves internal stresses and modifies microconstituents such as martensite/austenite (M/A) islands. The response variables were yield strength (Rp0.2), and Charpy V-notch impact energies at -40°C and -80°C (KV2), which are critical indicators for the performance of steel castings in low-temperature environments.
The material used in this study was a 10CrNiCu low-alloy steel casting, with a chemical composition as shown in Table 1. The as-cast microstructure consisted of ferrite, pearlite, and bainite, with pronounced segregation bands. These bands, enriched in alloying elements, exhibited higher hardenability, leading to the formation of granular bainite upon cooling, while the matrix primarily comprised ferrite and pearlite. The initial mechanical properties of the steel castings are summarized in Table 2, indicating a yield strength of approximately 436 MPa but relatively low impact toughness at sub-zero temperatures.
| C | Si | Mn | S | P | Cr+Ni+Cu | Fe |
|---|---|---|---|---|---|---|
| 0.11 | 0.2 | 1.8 | <0.05 | <0.05 | ≤3.0 | Bal. |
| Property | Rp0.2 (MPa) | Rm (MPa) | A (%) | Z (%) | -40°C KV2 (J) | -80°C KV2 (J) |
|---|---|---|---|---|---|---|
| Average Value | 436 | 561 | 26 | 65 | 64 | 9 |
The heat treatment process involved normalizing at 900°C followed by tempering at various temperatures and times. Normalizing was performed to austenitize the steel castings and promote grain refinement through phase transformation. Multiple normalizing cycles were investigated to assess their cumulative effect on microstructural homogenization. After normalizing, the samples were tempered to relieve stresses and decompose metastable phases. The orthogonal array L9(3^3) was used, resulting in nine distinct heat treatment combinations, as outlined in Table 3. This design ensures balanced sampling and allows for the analysis of each factor’s influence without the need for a full factorial experiment, which would require 27 runs. For each combination, mechanical tests were conducted according to ASTM standards, and microstructural analysis was performed using optical microscopy and transmission electron microscopy (TEM).
| Experiment No. | Normalizing Cycles | Tempering Temperature (°C) | Tempering Time (h) |
|---|---|---|---|
| 1 | 1 | 570 | 2 |
| 2 | 1 | 600 | 4 |
| 3 | 1 | 630 | 8 |
| 4 | 2 | 570 | 4 |
| 5 | 2 | 600 | 8 |
| 6 | 2 | 630 | 2 |
| 7 | 3 | 570 | 8 |
| 8 | 3 | 600 | 2 |
| 9 | 3 | 630 | 4 |
The results of the orthogonal experiment are presented in Table 4, which includes the measured mechanical properties for each treatment. To analyze the effects, I calculated the mean values for each factor level and performed analysis of variance (ANOVA). The influence of normalizing cycles, tempering temperature, and tempering time on yield strength and impact toughness is visualized in Figure 1, though the figure is not explicitly included here, its interpretation is discussed. For yield strength, the data showed that increasing normalizing cycles slightly enhanced strength, while higher tempering temperatures and longer tempering times significantly reduced it. This reduction can be attributed to the recovery of dislocations and the coarsening of precipitates, which soften the matrix of steel castings. The relationship between yield strength and grain size can be described by the Hall-Petch equation:
$$\sigma_y = \sigma_0 + k_y d^{-1/2}$$
where $\sigma_y$ is the yield strength, $\sigma_0$ is the friction stress, $k_y$ is the strengthening coefficient, and $d$ is the average grain diameter. In steel castings, normalizing refines grains, thereby increasing $\sigma_y$, but tempering counteracts this by reducing dislocation density.
| Exp. No. | Rp0.2 (MPa) | -40°C KV2 (J) | -80°C KV2 (J) |
|---|---|---|---|
| 1 | 508 | 119 | 19 |
| 2 | 476 | 128 | 46 |
| 3 | 420 | 180 | 106 |
| 4 | 508 | 95 | 59 |
| 5 | 454 | 162 | 24 |
| 6 | 452 | 162 | 83 |
| 7 | 478 | 174 | 20 |
| 8 | 511 | 187 | 49 |
| 9 | 445 | 151 | 43 |
For impact toughness, the -40°C and -80°C KV2 values generally improved with increasing tempering temperature, due to the decomposition of brittle M/A islands in the bainitic regions of the segregation bands. However, extending tempering time beyond 4 hours did not consistently enhance toughness, as prolonged exposure led to the precipitation and growth of carbides and Cu-rich particles, which can act as stress concentrators. The effect of normalizing cycles was nonlinear; two cycles provided the best toughness, likely because excessive normalizing might induce grain growth or other undesirable transformations in steel castings. The ANOVA results, summarized in Table 5, confirmed that tempering temperature had a highly significant effect on all properties, while normalizing cycles and tempering time were significant only for certain responses. This underscores the complexity of optimizing heat treatment for steel castings, where multiple factors interact.
| Property | Factor | Sum of Squares | Degrees of Freedom | Mean Square | F-Value | Significance |
|---|---|---|---|---|---|---|
| Rp0.2 | Normalizing Cycles | 236 | 2 | 118 | 2.62 | * |
| Tempering Temperature | 11347 | 2 | 5674 | 126 | *** | |
| Tempering Time | 4755 | 2 | 2377 | 53 | *** | |
| -40°C KV2 | Normalizing Cycles | 6355 | 2 | 3178 | 2.05 | * |
| Tempering Temperature | 8593 | 2 | 4297 | 2.78 | * | |
| Tempering Time | 5073 | 2 | 2537 | 1.64 | o | |
| -80°C KV2 | Normalizing Cycles | 289 | 2 | 145 | 0.07 | o |
| Tempering Temperature | 9847 | 2 | 4924 | 2.52 | * | |
| Tempering Time | 2089 | 2 | 1045 | 0.54 | o |
Microstructural analysis provided insights into these mechanical trends. After a single normalizing cycle, the grain size of the steel castings refined significantly from an initial ASTM grain size number of 7.5 to approximately 11, as shown in Figure 2a. With two and three cycles, further refinement occurred, reaching grain size numbers of 11.5 to 12 (Figure 2b and 2c). This refinement directly contributed to the increase in yield strength via the Hall-Petch mechanism. However, the diminishing returns after two cycles suggest that grain growth might begin to occur if normalizing is repeated excessively, which is a critical consideration for industrial heat treatment of steel castings.
The evolution of microstructure in the segregation bands was particularly revealing. In the normalized state, the bands contained granular bainite with M/A islands, as observed in TEM images (Figure 3a). After tempering at 570°C for 2 hours, these M/A islands remained partially retained, maintaining a lath morphology (Figure 3b). In contrast, tempering at 630°C for 8 hours led to complete recovery of the M/A islands, transforming them into fine polygonal ferrite with reduced dislocation density, and promoting the precipitation of spherical carbides and coarser Cu particles (Figure 3c). This microstructural change explains the drastic drop in yield strength and the improvement in toughness at higher tempering temperatures. The precipitation behavior can be modeled using the Lifshitz-Slyozov-Wagner theory for Ostwald ripening:
$$\bar{r}^3 – \bar{r}_0^3 = \frac{8\gamma D C_\infty V_m}{9RT}t$$
where $\bar{r}$ is the average precipitate radius, $\bar{r}_0$ is the initial radius, $\gamma$ is the interfacial energy, $D$ is the diffusion coefficient, $C_\infty$ is the solubility, $V_m$ is the molar volume, $R$ is the gas constant, $T$ is temperature, and $t$ is time. For steel castings containing Cu, prolonged tempering leads to growth of Cu precipitates, which can hinder dislocation motion but also embrittle the matrix if overgrown.
Based on the orthogonal analysis, I optimized the heat treatment for steel castings by selecting two normalizing cycles at 900°C, tempering at 630°C for 4 hours. This combination was predicted to yield a balance of properties, with an estimated yield strength of 440 MPa, -40°C KV2 of 178 J, and -80°C KV2 of 85 J, using the following estimation equation derived from orthogonal effects:
$$\hat{\mu} = \bar{\mu} + \hat{a}_2 + \hat{b}_3 + \hat{c}_2$$
where $\hat{\mu}$ is the estimated response, $\bar{\mu}$ is the overall mean, and $\hat{a}_2$, $\hat{b}_3$, $\hat{c}_2$ are the effect estimates for two normalizing cycles, 630°C tempering, and 4 hours tempering, respectively. Experimental validation confirmed these predictions, with actual values of Rp0.2 = 441 MPa, -40°C KV2 = 246 J, and -80°C KV2 = 137 J, demonstrating the effectiveness of the optimized process for steel castings.
In discussion, the findings highlight the importance of tailored heat treatment for steel castings. Normalizing cycles refine grains, enhancing strength and toughness, but excessive cycles offer minimal benefits. Tempering temperature is the most critical factor; higher temperatures promote recovery and dissolution of brittle phases, improving toughness but reducing strength. Tempering time should be optimized, as overly long times lead to precipitate coarsening, which can degrade toughness. These insights are crucial for manufacturers of steel castings, enabling them to adjust processes based on desired performance metrics. For instance, in applications requiring high toughness at low temperatures, such as Arctic offshore structures, the optimized treatment ensures that steel castings meet stringent standards.
Moreover, the orthogonal experimental design proved efficient for this optimization, reducing the number of trials while providing comprehensive data. This method can be extended to other grades of steel castings, incorporating additional factors like cooling rates or alloy modifications. Future work could explore the interactions between factors more deeply, perhaps using response surface methodology, to further refine heat treatment protocols for steel castings.
In conclusion, the optimization of heat treatment for 10CrNiCu low-alloy steel castings via orthogonal design has successfully identified a process that enhances low-temperature impact toughness without excessively compromising yield strength. The recommended protocol—two normalizing cycles at 900°C followed by tempering at 630°C for 4 hours—results in a yield strength of 440 MPa and significantly improved impact energies at -40°C and -80°C. This study underscores the value of systematic experimentation in advancing the performance of steel castings, ensuring their reliability in critical applications. As demand for high-performance steel castings grows, such optimization efforts will remain essential for pushing the boundaries of material science and engineering.