Steel casting quenching is a complex process that involves simultaneous temperature evolution, phase transformation, and stress–strain development. The selection of an appropriate quenching process directly determines whether the final mechanical properties of the steel casting meet the service requirements. Experimental determination of optimum steel casting quenching schedules is expensive and time consuming, and each experiment is only valid for a specific steel composition and a particular process window. Therefore, the use of computer numerical simulation to calculate the transient temperature field, to predict the quenched microstructure and hardness, and to evaluate the influence of different quenching media and process parameters on steel casting performance is of great practical significance. In this dissertation, a three-dimensional heat conduction model for the quenching of steel castings is established. The latent heat of phase transformation, temperature-dependent thermal physical properties, and the temperature-dependent interfacial heat transfer coefficients of various quench media are all considered. Based on the Maynier–Cacsi predictive model, a method for predicting the quenched microstructure and hardness of steel castings is proposed. A numerical solver based on the finite difference method is developed, and a dedicated software system for the simulation of the steel casting quenching process and the prediction of microstructure and mechanical properties is implemented. The system is applied to a stepped test block and to a crane brake wheel. The temperature fields and microstructure distributions are analyzed for water quenching, oil quenching, and water-to-oil quenching. The results show that the water-to-oil quenching process can provide a high surface hardness while preserving a tough core, which is beneficial for the comprehensive mechanical properties of the steel casting.
1. Introduction
Casting is one of the oldest metal forming technologies and remains an important production method for blanks in modern manufacturing. Steel castings account for about 15% of the total casting output in many industrial countries and are widely used in automotive, heavy machinery, aerospace, energy, and many other fields. For machine parts that require high comprehensive mechanical properties, steel castings are often the preferred choice. The final performance of a steel casting depends not only on its chemical composition and the casting process, but also on the subsequent heat treatment. Among all heat treatment procedures, quenching is the most critical step for increasing the surface hardness and wear resistance of steel castings. An improperly designed quenching process may lead to insufficient hardening, excessive distortion, or even cracking. Thus, a deep understanding of the quenching process of steel castings is essential for process optimization and quality control.
The traditional way to study the quenching behavior of steel castings is to perform physical experiments, in which process parameters are varied and the resulting microstructure and properties are measured. However, such experimental methods are usually blind, costly, and time consuming. Especially for large steel castings and new product development, the trial-and-error approach requires numerous iterations and long development cycles. With the rapid development of numerical simulation technology, computer simulation has become a powerful tool in the field of casting and heat treatment. Numerical simulation can provide an intuitive and quantitative description of the transient temperature distribution, the evolution of phases, and the final residual stress state. By combining temperature field simulation with empirical microstructure prediction models, it is possible to obtain the distribution of microstructure and hardness in a steel casting after quenching without performing extensive experiments. This enables engineers to test different quenching strategies, optimize process parameters, and reduce the cost and time of product development.
Extensive research has been carried out on the numerical simulation of quenching processes. In the early 1970s, Swedish researchers developed methods to predict continuous cooling transformation behavior from isothermal transformation diagrams using the additivity rule. Later, Maynier and co-workers analyzed a large number of experimentally measured continuous cooling transformation (CCT) diagrams and established empirical formulas that relate the critical cooling rates and the hardness of single-phase microstructures to the chemical composition and austenitizing conditions of the steel. Cacsi further modified the Maynier model and provided more accurate formulas for the critical cooling rates and hardness values. These models form the basis of many industrial prediction systems for the heat treatment of steel components.
In this thesis, the focus is placed on the numerical simulation of the temperature field during the quenching of steel castings and the subsequent prediction of the quenched microstructure and hardness. The finite difference method (FDM) is adopted to discretize the three-dimensional heat conduction equation. A dedicated computer program is developed to solve the temperature field and to compute the cooling rate at 700°C. The Maynier–Cacsi model is then used to predict the phase fractions and hardness distribution throughout the steel casting. The developed system is validated on a stepped test block and applied to a real crane brake wheel. The influence of different quenching media and different quenching procedures is discussed in detail. Throughout the thesis, the term steel casting is used to emphasize that the investigated objects are cast steel components with possible initial segregation, shrinkage porosity, and other casting defects that may influence the quenching response.
2. Mathematical Model of the Temperature Field during Steel Casting Quenching
2.1 Basic Heat Transfer Mechanisms
Heat transfer during the quenching of a steel casting involves three fundamental mechanisms: conduction, convection, and radiation. Inside the steel casting, heat is transferred by conduction. At the boundary between the steel casting and the quenching medium, heat is removed by convection and, at high temperatures, also by radiation. In most quenching operations, the boundary condition can be described by Newton’s law of cooling, which is a convective boundary condition of the third kind.
Based on Fourier’s law, the heat conduction within an isotropic material is expressed as:
$$q = -\lambda \frac{\partial T}{\partial n}$$
where \(q\) is the heat flux, \(\lambda\) is the thermal conductivity, \(T\) is the temperature, and \(n\) is the direction normal to the isothermal surface. For a three-dimensional domain with an internal heat source, the governing equation is:
$$\rho c \frac{\partial T}{\partial t} = \frac{\partial}{\partial x}\left(\lambda \frac{\partial T}{\partial x}\right) + \frac{\partial}{\partial y}\left(\lambda \frac{\partial T}{\partial y}\right) + \frac{\partial}{\partial z}\left(\lambda \frac{\partial T}{\partial z}\right) + \dot{Q}$$
where \(\rho\) is the density, \(c\) is the specific heat capacity, \(t\) is time, and \(\dot{Q}\) is the internal heat source per unit volume. In the quenching process of steel castings, the internal heat source mainly arises from the latent heat released during solid-state phase transformations, such as the decomposition of austenite into martensite, bainite, or ferrite–pearlite. The latent heat is treated by the equivalent heat source method, in which the latent heat rate is added to the energy equation at the temperature nodes where phase transformation occurs.
2.2 Initial and Boundary Conditions
For the steel casting quenching simulation, the initial condition is usually the uniform austenitizing temperature. If the steel casting is fully austenitized at temperature \(T_0\) before being transferred to the quench tank, the initial condition is:
$$T(x,y,z,0) = T_0$$
For induction-hardened steel castings, the initial temperature may be non-uniform:
$$T(x,y,z,0) = T_0(x,y,z)$$
Three types of boundary conditions are commonly used in heat transfer analysis. In the quenching process of steel castings, the third kind of boundary condition is the most appropriate:
$$-\lambda \left.\frac{\partial T}{\partial n}\right|_s = h \left(T_w – T_f\right)$$
where \(h\) is the convective heat transfer coefficient between the steel casting surface and the quenching medium, \(T_w\) is the surface temperature of the steel casting, and \(T_f\) is the temperature of the quenching fluid far away from the surface.
2.3 Temperature-Dependent Thermal Properties
The thermal physical properties of steel castings, particularly the thermal conductivity and the specific heat capacity, vary significantly with temperature during quenching. In the present model, the thermal conductivity and the specific heat capacity are treated as functions of temperature. For the ZG45 steel casting used in this thesis, the thermal conductivity data from 0°C to 900°C are listed in Table 1.
| T (°C) | 30 | 105 | 185 | 270 | 350 | 430 | 515 | 595 | 675 | 760 | 840 | 900 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| \(\lambda\) (W/(m·°C)) | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 27.7 |
The corresponding specific heat capacity values are given in Table 2.
| T (°C) | 30 | 100 | 200 | 300 | 400 | 450 | 500 | 550 | 700 | 750 | 850 | 900 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| \(c_p\) (J/(kg·°C)) | 455.6 | 476 | 500 | 518 | 535.2 | 543.4 | 552.4 | 560 | 583.5 | 590.8 | 607.2 | 614.2 |
Fitting the data in Table 1 and Table 2 yields the following functions:
$$\lambda(T) = 0.0123T + 16.697$$
$$c_p(T) = -5 \times 10^{-5} T^2 + 0.2259T + 453.04$$
where \(T\) is in °C, \(\lambda\) is in W/(m·°C), and \(c_p\) is in J/(kg·°C).
3. Numerical Solution Based on the Finite Difference Method
The finite difference method is adopted to solve the three-dimensional heat conduction equation. A steel casting is discretized into a set of cubic cells with uniform side length \(\Delta x\). Each cell exchanges heat with its six neighboring cells. For an interior cell \(i\), the discrete energy balance can be written as:
$$\rho_i c_{pi} \Delta x^3 \frac{T_i^{t+\Delta t} – T_i^t}{\Delta t} = \sum_{j=1}^{6} \frac{T_j^t – T_i^t}{\frac{\Delta x}{2\lambda_i} + \frac{\Delta x}{2\lambda_j}} \Delta x^2$$
After rearrangement, the temperature of cell \(i\) at the new time step is:
$$T_i^{t+\Delta t} = T_i^t + \frac{\Delta t}{\rho_i c_{pi} \Delta x^2} \sum_{j=1}^{6} \lambda_{ij} \left(T_j^t – T_i^t\right)$$
where \(\lambda_{ij}\) is the effective thermal conductivity between cell \(i\) and cell \(j\). For a boundary cell, the heat exchange with the quenching medium is described by:
$$\rho_i c_{pi} \Delta x^3 \frac{T_i^{t+\Delta t} – T_i^t}{\Delta t} = \sum_{j=1}^{m} \frac{T_j^t – T_i^t}{\frac{\Delta x}{2\lambda_i} + \frac{\Delta x}{2\lambda_j}} \Delta x^2 + \sum_{k=1}^{q} h_{ik} \left(T_f – T_i^t\right) \Delta x^2$$
where \(m\) is the number of internal neighboring cells, \(q\) is the number of boundary faces, and \(h_{ik}\) is the interfacial heat transfer coefficient between the boundary face and the quenching fluid. The stability condition requires that the time step satisfies:
$$\Delta t \le \frac{\rho c_p \Delta x^2}{6\lambda}$$
This condition ensures that the coefficients of the discretized equation remain non-negative and that the numerical solution is stable.
The interfacial heat transfer coefficient \(h\) is a strong function of the steel casting surface temperature. For the three quenching media considered in this work, namely water at 20°C, L-AN22 total loss system oil at 20°C, and 10% NaCl aqueous solution at 25°C, the heat transfer coefficient curves are fitted as piecewise polynomial functions. For example, the heat transfer coefficient of water is expressed as:
$$
h(T) = \begin{cases}
-8 \times 10^{-10} T^4 + 5 \times 10^{-7} T^3 – 9 \times 10^{-5} T^2 + 0.0067T, & 0 \le T < 300 \\
-0.0458 T^2 + 3.5705T + 1000.7, & 300 \le T < 900
\end{cases}
$$
where \(h\) is in W/(m\(^2\)·°C) and \(T\) is the surface temperature in °C.
The latent heat released during phase transformation is treated as an equivalent heat source. In each time step, the increment of phase fraction is computed from the transformation kinetics, and the corresponding latent heat is added to the energy balance:
$$\dot{Q} = \Delta H \frac{\Delta V}{\Delta t}$$
where \(\Delta H\) is the enthalpy change of the phase transformation and \(\Delta V\) is the increment of the transformed volume fraction during the time step \(\Delta t\).
4. Prediction of Microstructure and Hardness of Steel Castings
4.1 Continuous Cooling Transformation and Critical Cooling Rates
The final microstructure of a steel casting after quenching strongly depends on the cooling rate. According to the CCT diagram, different cooling rates lead to different phase constituents. In this work, the Maynier–Cacsi model is used to predict the critical cooling rates from the chemical composition and the austenitizing conditions. The fundamental formula is:
$$\log V = K – A W(C) – B W(Si) – D W(Mn) – E W(Ni) – F W(Cr) – G W(Mo) – H W(V)$$
where \(V\) is the critical cooling rate in °C/h, and \(W(X)\) is the weight percentage of element X. The coefficients \(K, A, B, D, E, F, G\) depend on the specific critical cooling rate. The Maynier model provides coefficients for the critical cooling rates for martensite formation, bainite formation, and ferrite–pearlite transformation. The Cacsi model extends the original Maynier model with additional critical cooling rates. Table 3 lists the coefficients proposed by Cacsi for various critical cooling rates.
| Coefficient | \(V_1\) | \(V_1^{90}\) | \(V_1^{50}\) | \(V_1^{10}\) | \(V_1^{0}\) | \(V_2^{0}\) | \(V_2^{10}\) | \(V_2^{50}\) | \(V_2^{90}\) | \(V_2\) |
|---|---|---|---|---|---|---|---|---|---|---|
| K | 9.81 | 8.76 | 8.20 | 9.80 | 8.56 | 10.55 | 9.06 | 8.04 | 8.40 | 8.56 |
| A | 4.62 | 4.04 | 3.00 | 3.90 | 1.50 | 4.80 | 4.11 | 3.40 | 2.80 | 1.50 |
| B | 0.78 | 0.86 | 0.79 | -0.54Mn+2.45√Mn | 1.84 | 0.80 | 0.90 | 1.15 | 1.51 | 1.84 |
| D | 0.41 | 0.36 | 0.57 | 0.46 | 0.70 | 0.72 | 0.60 | 0.96 | 1.03 | 0.78 |
| E | 0.80 | 0.58 | 0.67 | 0.50 | 1.24 | 1.07 | 1.00 | 1.00 | 1.10 | 1.24 |
| F | 0.66 | 0.97 | 0.94 | 1.16 | 1.46 | 1.58 | 2.00 | 2.00 | 2.31 | 2√Mo |
| G | 0.0018 | 0.001 | 0.0012 | 0.002 | 0.002 | 0.0026 | 0.0013 | 0.007 | 0.0014 | 0.002 |
The austenitizing parameter \(Pa\) is used to account for the effect of the austenitizing temperature and holding time:
$$Pa = \frac{1}{R}\left(\frac{1}{T} – \frac{\ln t}{\Delta H}\right)$$
where \(T\) is the austenitizing temperature, \(t\) is the holding time, \(R\) is the ideal gas constant, and \(\Delta H\) is the activation energy (taken as 110000 J/mol).
4.2 Hardness of Single-Phase Microstructures
The Maynier–Cacsi model also provides formulas for the hardness of martensite (\(H_M\)), bainite (\(H_B\)), and ferrite–pearlite (\(H_{F+P}\)). The hardness values are expressed in Vickers hardness (HV). The formulas used in this thesis are:
$$H_M = 121.156 + 902.6 W(C) + 26.68 \log V$$
$$H_B = 323 + 185 W(C) + 330 W(Si) + 153 W(Mn) + 65 W(Ni) + 144 W(Cr) + 191 W(Mo) + \left(89 + 53 W(C) – 55 W(Si) – 22 W(Mn) – 10 W(Ni) – 20 W(Cr) – 33 W(Mo)\right) \log V$$
$$H_{F+P} = -437 + 3300 W(C) – 5343 W(C)^2 + \left(1329 W(C) – 744 W(C)^2 – 4 W(Ni) + 15 W(Cr) + 135.4\right)\log V$$
where \(V\) is the cooling rate at 700°C expressed in °C/h. The cooling rate at 700°C is obtained from the numerical temperature field simulation. For each finite difference cell, the cooling rate is calculated by:
$$V_{700} = \frac{T_1 – T_2}{\Delta t}$$
where \(T_1 > 700^{\circ}\mathrm{C}\) at time \(t\) and \(T_2 < 700^{\circ}\mathrm{C}\) at time \(t+\Delta t\), respectively.
4.3 Prediction of Phase Fractions and Overall Hardness
Once the actual cooling rate at 700°C is determined for each cell of the steel casting, it is compared with the critical cooling rates calculated from the Cacsi model. The phase fractions are then obtained by linear interpolation between the appropriate critical cooling rates. For example, if the cooling rate \(V\) lies between \(V_1\) and \(V_1^{50}\), the volume fraction of martensite is estimated as:
$$W_M = W_{M,H} – \left(W_{M,H} – W_{M,L}\right)\frac{V – V_L}{V_H – V_L}$$
where \(V_H\) and \(V_L\) are the two bounding critical cooling rates, and \(W_{M,H}\) and \(W_{M,L}\) are the corresponding martensite fractions. Similar interpolation procedures are used for bainite and ferrite–pearlite.
The overall hardness of the steel casting after quenching is calculated by the rule of mixtures:
$$HV = W_M H_M + W_B H_B + W_{F+P} H_{F+P}$$
where \(W_M\), \(W_B\), and \(W_{F+P}\) are the volume fractions of martensite, bainite, and ferrite–pearlite, respectively, and \(H_M\), \(H_B\), and \(H_{F+P}\) are the corresponding single-phase hardness values.
5. Development of the Simulation System for Steel Casting Quenching
A dedicated software system is developed in the Visual Studio 2005 environment. The system integrates three major functional modules: temperature field calculation, microstructure prediction, and hardness prediction. The overall design is illustrated in the flow chart of the system architecture. The temperature field module reads the finite difference mesh of the steel casting, the thermophysical property data, the quenching medium heat transfer coefficients, and the process parameters. After solving the transient temperature field, the module outputs the temperature history and the cooling rate at 700°C for every cell of the steel casting. The microstructure prediction module takes the chemical composition of the steel casting, the austenitizing conditions, and the cooling rate file as input. It calculates the critical cooling rates, compares them with the local cooling rate, and outputs the phase fractions. The hardness prediction module combines the phase fractions with the single-phase hardness formulas and outputs the hardness distribution.
The user interface of the system contains a main function selection window. The user can choose to run the temperature field simulation, the microstructure prediction, or the hardness prediction. Parameter input dialogs allow the user to enter the thermal conductivity, specific heat capacity, density, heat transfer coefficient, initial temperature, austenitizing temperature, austenitizing time, quenching medium type, and medium temperature. The system supports both constant and temperature-dependent material properties. The temperature-dependent data are entered as a table, and the software performs linear interpolation between the data points during the simulation. After the calculation is finished, the results are saved to binary files that can be post-processed by third-party visualization tools.
The numerical solver is carefully verified against the stability condition. The time step is automatically adjusted according to the cell size and the thermal diffusivity of the steel casting. The latent heat release is evaluated every time step using the phase transformation kinetics. The system has been applied to several steel casting components, and the simulation results are in good agreement with the expected quenching behavior.
6. Application and Case Studies
6.1 Stepped Test Block
To investigate the influence of different quenching media on the temperature field and the final microstructure of steel castings, a stepped test block is used. The three-dimensional model of the stepped block consists of five steps with different thicknesses, representing different sections of a real steel casting. The material is ZG45, with the chemical composition listed in Table 4.
| C | Si | Mn | S | P | Cr | Ni | Mo | Cu | V |
| ≤0.50 | ≤0.60 | ≤0.90 | ≤0.04 | ≤0.04 | ≤0.35 | ≤0.30 | ≤0.20 | ≤0.30 | ≤0.05 |
The stepped block is initially austenitized at 860°C and then quenched in three different media: water at 20°C, L-AN22 oil at 20°C, and 10% NaCl aqueous solution at 25°C. The temperature field is computed at different times. Figure 1 shows a representative steel casting used in the manufacturing industry, illustrating the type of component that can be analyzed with the present system.
Thin sections of the stepped block cool much faster than thick sections. For example, after 15 s of water quenching, the surface temperature of the thinnest step drops below 200°C, while the thickest section still maintains a temperature above 500°C. The cooling curves at the center of each step are recorded. Figure 2 shows the cooling curves of five characteristic points located at the centers of the five steps. The five points are denoted as point 1 (thinnest) to point 5 (thickest). As expected, point 1 has the highest cooling rate and point 5 has the lowest cooling rate. A comparison of the cooling curves under the three media is made. Water produces the fastest cooling, oil produces the slowest cooling, and the 10% NaCl solution lies between them. In the temperature range above 200°C, the NaCl solution cools faster than oil, but below 200°C, oil has a slightly higher heat transfer coefficient than the NaCl solution, which is consistent with experimental observations reported in the literature.
Using the temperature field results, the cooling rate at 700°C is calculated for every cell of the stepped block. The microstructure prediction module is then applied. The predicted phase fractions of martensite, bainite, and ferrite–pearlite, as well as the hardness distribution, are shown in the output files. For water quenching, the surface and thin sections are almost fully martensitic with a Vickers hardness of about 643 HV. For oil quenching, a significant amount of bainite is present, and the hardness of the thin sections is lower. Table 5 summarizes the predicted phase fractions and hardness values at the five characteristic points for the three quenching media.
| Quenching medium | Point | Martensite (%) | Bainite (%) | Ferrite–pearlite (%) | Hardness (HV) |
|---|---|---|---|---|---|
| Water at 20°C | 1 | 41.8 | 58.2 | 0.0 | 434.7 |
| 2 | 55.2 | 44.8 | 0.0 | 483.9 | |
| 3 | 96.1 | 3.9 | 0.0 | 629.3 | |
| 4 | 100.0 | 0.0 | 0.0 | 642.9 | |
| 5 | 100.0 | 0.0 | 0.0 | 642.9 | |
| L-AN22 oil at 20°C | 1 | 24.3 | 75.7 | 0.0 | 374.9 |
| 2 | 29.6 | 70.4 | 0.0 | 393.2 | |
| 3 | 51.7 | 48.3 | 0.0 | 471.5 | |
| 4 | 92.0 | 8.0 | 0.0 | 614.7 | |
| 5 | 100.0 | 0.0 | 0.0 | 642.9 | |
| 10% NaCl at 25°C | 1 | 34.0 | 66.0 | 0.0 | 409.2 |
| 2 | 44.2 | 55.8 | 0.0 | 444.8 | |
| 3 | 76.5 | 23.5 | 0.0 | 559.5 | |
| 4 | 100.0 | 0.0 | 0.0 | 642.9 | |
| 5 | 100.0 | 0.0 | 0.0 | 642.9 |
The results confirm that water quenching provides the highest cooling capacity and the highest martensite fraction. Oil quenching results in lower hardness and more bainite. The 10% NaCl solution shows an intermediate cooling capacity. No ferrite–pearlite transformation is predicted in any case for the stepped steel casting because the cooling rates are all high enough to avoid the pearlite nose.
6.2 Crane Brake Wheel
The developed system is further applied to a real crane brake wheel made of ZG45 steel casting. The brake wheel has a diameter of 630 mm and a length of 260 mm. Six characteristic points are selected on the wheel: A on the outer surface, B at the center, C at the rim edge, D in the web, E at the hub center, and F near the hub surface. The finite difference mesh is generated with a uniform cell size of 3 mm, resulting in approximately 7 million cells.
Three quenching procedures are simulated: (1) water quenching from 860°C in 20°C water, (2) oil quenching in 20°C oil, and (3) water-to-oil quenching, in which the brake wheel is first quenched in water until the surface temperature drops to about 200°C and then transferred to oil for further cooling. The transient temperature fields are calculated at several instants. For water quenching, the surface temperature decreases very rapidly; after 5 s the surface temperature is about 600°C, and after 30 s the surface is below 200°C. The temperature distribution is highly non-uniform, with the core still above 600°C when the surface has reached room temperature. This severe temperature gradient may lead to large thermal stresses and quenching cracks. In oil quenching, the cooling is much slower; after 30 s the surface temperature is still above 550°C. The water-to-oil quenching procedure initially follows the water cooling curve, but after the transfer to oil, the cooling rate is reduced significantly, leading to a more uniform final temperature distribution.
The microstructure and hardness distributions of the brake wheel after quenching are predicted using the same procedure. Table 6 lists the phase fractions and hardness at the six characteristic points for the three quenching procedures.
| Quenching process | Point | Martensite (%) | Bainite (%) | Ferrite–pearlite (%) | Hardness (HV) |
|---|---|---|---|---|---|
| Water quenching | A | 100.0 | 0.0 | 0.0 | 642.9 |
| B | 68.4 | 31.6 | 0.0 | 530.9 | |
| C | 100.0 | 0.0 | 0.0 | 642.9 | |
| D | 71.4 | 28.7 | 0.0 | 541.3 | |
| E | 53.8 | 36.2 | 0.0 | 479.1 | |
| F | 75.8 | 24.2 | 0.0 | 557.2 | |
| Oil quenching | A | 96.1 | 3.1 | 0.0 | 629.0 |
| B | 40.4 | 49.6 | 0.0 | 431.2 | |
| C | 73.8 | 26.2 | 0.0 | 547.3 | |
| D | 42.0 | 38.1 | 0.0 | 437.1 | |
| E | 34.6 | 65.4 | 0.0 | 411.0 | |
| F | 46.6 | 53.4 | 0.0 | 453.5 | |
| Water-to-oil quenching | A | 100.0 | 0.0 | 0.0 | 642.9 |
| B | 50.0 | 50.0 | 0.0 | 465.6 | |
| C | 100.0 | 0.0 | 0.0 | 642.9 | |
| D | 52.0 | 48.0 | 0.0 | 472.7 | |
| E | 49.4 | 50.6 | 0.0 | 463.5 | |
| F | 57.0 | 43.0 | 0.0 | 490.3 |
Water quenching yields a fully martensitic surface with hardness above 640 HV, which is excellent for wear resistance. However, the core contains only about 54% martensite and 36% bainite, which may not provide sufficient toughness to withstand impact loads. Oil quenching produces a much softer surface, with point A having only 96% martensite and the core having only about 35% martensite. The surface hardness is still fairly high, but the wear resistance is lower than that of water-quenched steel castings. The water-to-oil quenching procedure produces a fully martensitic surface and a core with approximately 50% martensite and 50% bainite. The surface hardness is as high as 643 HV, while the core hardness is around 465 HV. This combination provides both high wear resistance and improved toughness, making the water-to-oil quenching process the most suitable procedure for the crane brake wheel.
The numerical simulation of the temperature field during steel casting quenching provides detailed information about the cooling behavior at every location of the component. The subsequent microstructure and hardness prediction allows engineers to evaluate the effect of different quenching processes without performing expensive physical experiments. The methodology presented in this thesis is therefore a valuable tool for the optimization of quenching processes for steel castings.
7. Conclusion and Outlook
In this thesis, a complete numerical simulation and prediction system for the quenching of steel castings has been developed. The main conclusions can be summarized as follows.
First, the heat transfer during steel casting quenching is governed by a nonlinear three-dimensional heat conduction equation with temperature-dependent thermophysical properties and latent heat release. The finite difference method provides a stable and efficient solution for the transient temperature field when the time step is chosen according to the stability criterion. The interfacial heat transfer coefficient is the most critical parameter for the accuracy of the simulation. The temperature-dependent heat transfer coefficients for water, oil, and salt solution significantly affect the cooling curves and the final microstructure of the steel casting.
Second, the Maynier–Cacsi model is a practical and robust approach for predicting the quenching response of steel castings. By combining the numerically computed cooling rate at 700°C with the critical cooling rates obtained from the chemical composition, the phase fractions of martensite, bainite, and ferrite–pearlite can be predicted. The hardness distribution is then calculated using the rule of mixtures. The predicted results for the stepped steel casting block show that water quenching provides the highest martensite content, followed by the NaCl solution, while oil quenching gives the lowest martensite content. The simulation results are in qualitative agreement with the cooling capacities of the three media.
Third, the application to the crane brake wheel demonstrates the ability of the system to handle a real industrial steel casting with a complex geometry. The water-to-oil quenching process is found to be the optimal procedure for the brake wheel because it produces a hard martensitic surface and a tough bainitic core. This conclusion is consistent with the expected performance requirements of the component.
Future work should focus on the following aspects. More accurate and comprehensive heat transfer coefficient data for various quenching media and agitation conditions should be collected and integrated into the system. The initial microstructure of the steel casting before quenching, which depends on the preceding casting and cooling process, should be taken into account to achieve a fully integrated simulation of the entire manufacturing chain. In addition, the stress and strain fields during quenching should be coupled with the temperature and microstructure fields to predict distortion and cracking. Finally, the pre-processing and post-processing capabilities of the system should be improved to provide a more user-friendly environment for industrial applications.
In summary, the developed system offers a practical and efficient tool for the numerical simulation and prediction of microstructure and performance during steel casting quenching. It can effectively guide the design of quenching processes, reduce the need for expensive and time-consuming experiments, and ultimately improve the quality and reliability of steel castings used in various industrial applications.