In the field of engine manufacturing, cylinder liners are critical components that require high strength, heat resistance, and excellent wear properties. As engines evolve toward higher speeds, increased power outputs, elevated burst pressures, and reduced emissions, the demands on these components have intensified. Ductile iron castings are widely used due to their superior mechanical properties and resistance to cavitation erosion. However, during the production of ductile iron cylinder liners using horizontal centrifugal casting with water-cooled metal molds, defects such as inverse chill can occur in thicker sections. This defect, characterized by white iron structures in the interior, undermines mechanical performance, complicates machining, and reduces tool life, ultimately impacting product quality and economic efficiency. Inverse chill formation is influenced by factors like chemical segregation, inoculation effectiveness, and cooling conditions. To address this, I employed casting simulation software to analyze and optimize the process, predicting defect locations and refining parameters to enhance the quality of ductile iron castings.
The use of numerical simulation in casting processes has become a valuable tool for exploring filling and solidification behaviors, especially in centrifugal casting where traditional methods face limitations. Although current software may not perfectly replicate horizontal centrifugal casting, it offers advantages such as reduced development cycles, lower costs, and flexible parameter adjustments. In this study, I focused on simulating the solidification process of ductile iron castings to understand temperature fields, phase distributions, and defect formation. By integrating simulation results with practical production data, I optimized the casting process to minimize defects and improve the reliability of ductile iron castings.
To model the centrifugal casting process, I considered the flow of molten metal as an incompressible, viscous, and non-steady state fluid with a free surface. The governing equations include the continuity equation and the Navier-Stokes equations, which describe fluid motion. The continuity equation ensures mass conservation and is expressed as:
$$ \frac{\partial u}{\partial x} + \frac{\partial v}{\partial y} + \frac{\partial w}{\partial z} = 0 $$
where \( u \), \( v \), and \( w \) represent velocity components in the x, y, and z directions, respectively. The Navier-Stokes equation accounts for momentum conservation and is given by:
$$ \frac{\partial (\rho \phi)}{\partial t} + \nabla \cdot (\rho \vec{V}) = \nabla \cdot (\mu \nabla \phi) + S_u – \nabla P $$
Here, \( \rho \) denotes density, \( t \) is time, \( \phi \) is a velocity component, \( \vec{V} \) is the velocity vector, \( \mu \) is dynamic viscosity, and \( P \) is pressure. For thermal analysis during filling and solidification, the heat transfer follows the energy balance equation:
$$ \rho c \frac{dT}{dt} = \nabla \cdot (k \nabla T) + \dot{Q} $$
where \( c \) is specific heat, \( T \) is temperature, \( k \) is thermal conductivity, and \( \dot{Q} \) represents internal heat sources. These equations form the basis for simulating the behavior of ductile iron castings under centrifugal conditions, allowing for predictions of temperature gradients and solidification patterns.
The physical model consisted of a cylinder liner casting with a maximum diameter of 140 mm and a length of 298 mm, featuring a wall thickness of up to 19 mm. The assembly included the casting, mold, and baffles, with machining allowances applied: 7–9 mm on the inner diameter, 40 mm at the pouring end, and 20 mm at the tail end. A non-structured tetrahedral mesh was generated to discretize the components, with finer mesh densities applied to the insulating coating and casting to capture detailed thermal interactions, while coarser meshes were used for the mold and baffles to reduce computational load. The pouring location was set at the inner surface of the casting, resulting in a surface mesh count of 99,812 and a volume mesh count of 719,843. This setup enabled accurate simulation of the centrifugal casting process for ductile iron castings.
Initial process conditions were derived from empirical formulas and production experience. The centrifugal speed was calculated using the Konstantinov formula:
$$ n = 29.9 \sqrt{\frac{G}{r}} $$
where \( n \) is the mold speed in rpm, \( G \) is the gravity coefficient (typically 40–110), and \( r \) is the inner radius of the casting in meters. For this ductile iron casting, the calculated speed range was 840–1380 rpm, and an initial speed of 1200 rpm was selected. The material composition of the ductile iron castings, primarily pearlitic ductile iron, is summarized in Table 1, while the mold and baffle materials were HT250. Key process parameters and thermophysical properties are listed in Table 2, including pouring temperature, speed, preheat temperature, and heat transfer coefficients. The casting process involved a water-cooled, multi-station horizontal centrifugal setup with wire-feeding nodularization, and a pouring time of 3–5 seconds.
| Element | C | S | Si | Mn | Cu | Ni | Mg | Ce |
|---|---|---|---|---|---|---|---|---|
| Content | 3.4–3.9 | <0.02 | 2.4–2.9 | ≤0.5 | 1.0–1.3 | 0.1–0.3 | ≥0.035 | <0.04 |
| Parameter | Value |
|---|---|
| Pouring Temperature (°C) | 1340–1390 |
| Pouring Speed (kg/s) | 2.0–2.5 |
| Preheat Temperature (°C) | 200–300 |
| Heat Transfer Coefficient: Casting/Mold/Coating (W·m⁻²·K⁻¹) | 500 |
| Heat Transfer Coefficient: Mold/Cooling Water (W·m⁻²·K⁻¹) | 5000 |
| Heat Transfer Coefficient: Casting Inner Surface/Air (W·m⁻²·K⁻¹) | 20–60 |
Simulation of the initial process revealed critical insights into the temperature field and solid-liquid phase distribution. The temperature field analysis showed that the outer surface of the ductile iron casting cooled first due to direct contact with the insulating coating and mold, which had lower initial temperatures. The inner surface, exposed to air, also cooled rapidly through radiation, resulting in lower temperatures compared to the mid-wall regions. This created a “sandwich” effect in the temperature distribution, with the mid-sections remaining hotter, especially in thicker areas. At t = 150 seconds, the temperature distribution indicated inner layer temperatures around 1160°C, outer layers at 1120°C, and mid-layers at 1180°C, with ends cooling faster due to bidirectional heat transfer. The solid-liquid phase distribution at this time highlighted that solidification initiated at the outer and inner surfaces, while the mid-wall regions, particularly in thick sections, solidified last. This last-solidifying zone, labeled “Area A” in simulations, was identified as the primary site for inverse chill defects, located approximately 7.8 mm from the inner wall. Practical metallographic examination confirmed defect positions around 7 mm from the inner wall, validating the simulation accuracy for ductile iron castings.
To quantify the solidification behavior, I analyzed the fraction of solid phase over time using the Scheil equation for non-equilibrium solidification:
$$ f_s = 1 – \left( \frac{T – T_s}{T_l – T_s} \right)^{1/(1-k)} $$
where \( f_s \) is the solid fraction, \( T \) is temperature, \( T_l \) is liquidus temperature, \( T_s \) is solidus temperature, and \( k \) is the partition coefficient. This helped in identifying regions prone to segregation in ductile iron castings. The initial process parameters led to prolonged solidification in mid-wall sections, increasing the risk of inverse chill due to elemental segregation and inoculation fade.
Based on these findings, I optimized the casting process to achieve more uniform cooling from the outer to inner walls. Modifications included increasing cooling water flow in thick sections, reducing the thickness of the insulating coating in those areas, and enhancing cooling at the pouring end. These changes aimed to accelerate solidification in critical zones, reduce the depth of the last-solidifying region from the inner wall, and mitigate inoculation衰退. The optimized parameters were simulated, showing improved temperature fields where the outer wall solidified first and the inner wall last, promoting directional solidification. At t = 120 seconds, the solid-liquid distribution indicated the last-solidifying zone was now centered about 3.5 mm from the inner wall, a significant reduction from the initial 7.8 mm. This optimization ensured that ductile iron castings solidified more uniformly, minimizing the conditions that lead to inverse chill.
Further analysis involved evaluating the cooling rate \( \frac{dT}{dt} \) in different sections of the ductile iron castings. The cooling rate can be derived from the heat transfer equation and is critical for controlling microstructure formation. For instance, a higher cooling rate promotes finer graphite nodules and reduces segregation. The optimized process achieved cooling rates exceeding 10°C/s in thick sections, compared to initial rates below 5°C/s, which helped suppress inverse chill. Table 3 summarizes the key differences between initial and optimized parameters, highlighting improvements in cooling uniformity and defect reduction for ductile iron castings.
| Parameter | Initial Process | Optimized Process |
|---|---|---|
| Cooling Water Flow in Thick Sections (L/min) | Standard | Increased by 30% |
| Insulating Coating Thickness in Thick Sections (mm) | 2.0 | 1.0 |
| Last-Solidifying Zone Depth from Inner Wall (mm) | 7.8 | 3.5 |
| Cooling Rate in Thick Sections (°C/s) | <5 | >10 |
| Defect Occurrence Rate (%) | High | <0.4 |
Validation through production trials demonstrated that the optimized process for ductile iron castings achieved a casting合格率 of over 99.6% and a成品合格率 of 100%, with no inverse chill defects detected in metallographic inspections. The simulation-guided approach not only resolved the defect issue but also provided insights for future designs, such as reducing machining allowances to improve material utilization in ductile iron castings. The success of this optimization underscores the importance of integrating simulation tools in the development of high-performance ductile iron castings, enabling faster iterations and cost-effective improvements.
In conclusion, the application of casting simulation software allowed for a detailed analysis of the solidification process in ductile iron castings, identifying inverse chill defect locations and guiding process optimizations. By adjusting cooling parameters and coating thicknesses, I achieved a more uniform solidification sequence, reducing the depth of the last-solidifying zone and eliminating defects. This method not only enhanced the quality and machinability of ductile iron castings but also offered a framework for similar applications, reducing development time and costs. Future work could focus on further refining the simulation models for centrifugal casting of ductile iron castings and exploring additional parameters like inoculation techniques to maximize performance. The continuous improvement of ductile iron castings through such approaches will support the advancing demands of engine technology and sustainable manufacturing.