In the field of engine manufacturing, cylinder liners serve as critical components, directly influencing engine performance, durability, and emissions. With advancements toward higher speeds, increased power, elevated burst pressures, and reduced emissions, the demands on cylinder liner materials have intensified. Nodular cast iron, known for its excellent mechanical properties and resistance to cavitation erosion, has been widely adopted in this application. However, during production via horizontal centrifugal casting with water-cooled metal molds, a defect known as inverse chill or “反白口” often appears in thick-walled sections. This defect manifests as white iron structures within the matrix, compromising mechanical strength, machinability, and tool life, thereby affecting product quality and economic efficiency. The formation of inverse chill is associated with factors such as chemical segregation, inoculation effectiveness, and cooling conditions. To address this, we employed casting simulation software to analyze and optimize the process, reducing trial cycles and enhancing quality. This article details our approach, from initial simulation to optimized practice, emphasizing the role of simulation in advancing nodular cast iron casting technology.
The use of computer numerical simulation has revolutionized casting process design, offering insights into filling and solidification patterns without costly physical trials. Although simulation of horizontal centrifugal casting remains challenging due to complex fluid dynamics and heat transfer, it provides a valuable tool for predicting defects and guiding improvements. Our study focuses on a nodular cast iron cylinder liner produced by centrifugal casting, where inverse chill defects were observed. By simulating the solidification process, we aimed to identify defect-prone areas and optimize cooling conditions to achieve uniform solidification, thereby eliminating defects. This approach not only saves time but also enhances our understanding of centrifugal casting behavior for nodular cast iron components.
To begin, we established a mathematical model for the centrifugal casting process. The filling phase involves free-surface, viscous, incompressible, and unsteady flow, governed by the continuity equation and Navier-Stokes equations. For temperature field simulation during solidification, heat conduction must be considered. The key equations are as follows:
Continuity equation: $$ \frac{\partial u}{\partial x} + \frac{\partial v}{\partial y} + \frac{\partial w}{\partial z} = 0 $$ where \( u, v, w \) are velocity components.
Navier-Stokes equation: $$ \frac{\partial (\rho \phi)}{\partial t} + \nabla \cdot (\rho \vec{V}) = \nabla \cdot (\mu \nabla \phi) + S_u – \nabla P $$ where \( \rho \) is density, \( t \) is time, \( \phi \) is velocity component, \( \vec{V} \) is velocity vector, \( \mu \) is dynamic viscosity, and \( P \) is pressure.
Heat 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 \( \dot{Q} \) is internal heat source.
These equations form the basis for simulating the flow and solidification of molten nodular cast iron in centrifugal casting. We applied them to our specific geometry and process parameters to predict temperature fields and phase distributions.
The physical model represents a cylinder liner with a maximum diameter of 140 mm, length of 298 mm, and maximum wall thickness of 19 mm. The casting includes machining allowances: 7–9 mm on the inner diameter, 40 mm at the pouring end, and 20 mm at the tail end. The assembly consists of the mold, baffles, casting, and insulating coating. For mesh generation, we used unstructured tetrahedral grids, refining the mesh density at the coating and casting interfaces to capture thermal gradients accurately, while coarsening the mold and baffle grids to reduce computational load. The pouring gate was set at the inner surface, resulting in 99,812 face elements and 719,843 volume elements. This discretization enabled detailed analysis of heat transfer and solidification.
Initial process conditions were derived from production experience. The centrifugal speed was calculated using the Konstantinov empirical formula: $$ n = 29.9 \sqrt{\frac{G}{r}} $$ where \( n \) is mold speed in rpm, \( G \) is gravity coefficient (typically 40–110), and \( r \) is inner radius of the casting in meters. For our nodular cast iron liner, this yielded a speed range of 840–1380 rpm, and we selected 1200 rpm based on practical considerations. Other parameters included a pouring temperature of 1340–1390°C, pouring rate of 2.0–2.5 kg/s, mold preheat temperature of 200–300°C, and heat transfer coefficients as listed in Table 1. The chemical composition of the nodular cast iron is critical for properties; Table 2 summarizes the elemental ranges.
| Parameter | Value |
|---|---|
| Pouring Temperature (°C) | 1340–1390 |
| Pouring Speed (kg/s) | 2.0–2.5 |
| Mold Preheat Temperature (°C) | 200–300 |
| Heat Transfer Coefficient: Casting/Coating/Mold (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 |
| Element | Content |
|---|---|
| C | 3.4–3.9 |
| Si | 2.4–2.9 |
| Mn | ≤0.5 |
| Cu | 1.0–1.3 |
| Ni | 0.1–0.3 |
| Mg | ≥0.035 |
| Ce | <0.04 |
| S | <0.02 |
Simulation of the initial process revealed key insights into solidification behavior. The temperature field showed that the outer surface of the nodular cast iron casting cooled fastest due to contact with the insulating coating and mold, while the inner surface, exposed to air, cooled slower but still formed a temperature gradient. At t = 150 s, the temperature distribution exhibited a “sandwich” pattern: inner layer at 1160°C, outer layer at 1120°C, and middle layer at 1180°C, with ends cooler due to bidirectional heat transfer. This non-uniform cooling promoted last-solidification zones in thick sections. The solid-liquid phase distribution at t = 150 s indicated that solidification initiated at the outer and inner surfaces, with the last liquid pool located approximately 7.8 mm from the inner wall in thick areas (denoted as Zone A). This correlated with actual defect sites observed in production, where inverse chill appeared about 7 mm from the inner wall, confirming the simulation’s predictive accuracy for nodular cast iron defects.
To optimize the process, we targeted the cooling sequence. Inverse chill in nodular cast iron can arise from element segregation, inoculation fading, or rapid cooling in last-solidification regions. Based on our simulation, we hypothesized that enhancing cooling uniformity from outer to inner walls would mitigate defects. Specifically, we increased cooling water flow in thick sections, reduced insulation coating thickness there, and improved cooling at the pouring end. These adjustments aimed to accelerate solidification in critical zones, shorten last-solidification time, and minimize inoculation fading effects. The optimized parameters are summarized in Table 3, compared to initial values.
| Parameter | Initial Process | Optimized Process |
|---|---|---|
| Cooling Water Flow in Thick Sections | Standard | Increased by 30% |
| Insulation Coating Thickness in Thick Sections | Standard (1.0 mm) | Reduced to 0.5 mm |
| Cooling at Pouring End | Moderate | Enhanced via additional channels |
| Centrifugal Speed (rpm) | 1200 | 1200 (unchanged) |
| Pouring Temperature (°C) | 1340–1390 | 1350–1380 (tightened range) |
Re-simulation with optimized settings showed improved solidification patterns. The temperature field became more uniform, with the outer wall solidifying first and the inner wall last, achieving directional solidification. At t = 120 s, the solid-liquid distribution indicated the last-solidification zone shifted to about 3.5 mm from the inner wall, significantly closer than before. This reduction in depth of the last-solidification region decreased the risk of inverse chill formation in nodular cast iron. We validated these results through production trials, where the optimized process yielded castings with over 99.6% internal soundness and 100% final product qualification, aligning with simulation predictions. The success underscores the value of simulation in refining nodular cast iron casting processes.
Further analysis involved quantifying cooling rates and solidification times. We derived a simplified model for cooling rate \( R \) in centrifugal casting: $$ R = \frac{T_p – T_s}{t_f} $$ where \( T_p \) is pouring temperature, \( T_s \) is solidus temperature, and \( t_f \) is local solidification time. For nodular cast iron, typical values are \( T_p \approx 1370°C \) and \( T_s \approx 1150°C \). In the initial process, \( t_f \) in thick zones was prolonged due to insulation, leading to \( R \approx 5°C/s \). After optimization, increased cooling reduced \( t_f \), raising \( R \) to approximately 8°C/s, which suppressed carbide formation and promoted graphitization. This aligns with theories that faster cooling in last-solidification areas reduces inverse chill susceptibility in nodular cast iron.
The benefits of simulation extend beyond defect elimination. By understanding solidification dynamics, we can reduce machining allowances, improving material utilization. For instance, the original allowance of 7–9 mm on the inner diameter was based on conservative design, but simulation showed that with optimized cooling, sound casting could be achieved with 5–7 mm allowance, saving nodular cast iron material. Additionally, simulation allows rapid iteration of process variables without physical trials, cutting development time by an estimated 50%. This efficiency is crucial for meeting industry demands for high-performance nodular cast iron components.
In conclusion, our study demonstrates the effectiveness of casting simulation in optimizing the centrifugal casting process for nodular cast iron cylinder liners. Through numerical analysis, we predicted inverse chill locations and implemented targeted cooling modifications, resulting in defect-free production. The optimized process enhances cooling uniformity, shortens last-solidification times, and improves the overall quality of nodular cast iron castings. This approach not only resolves specific defects but also provides a framework for designing and refining casting processes for complex nodular cast iron parts. Future work could explore advanced inoculation techniques or real-time monitoring to further enhance nodular cast iron properties. Ultimately, simulation technology is a powerful ally in advancing the reliability and efficiency of nodular cast iron manufacturing, paving the way for more sustainable and high-performance engine components.
Reflecting on this experience, we recognize that nodular cast iron remains a versatile material for demanding applications, and its performance hinges on precise process control. By integrating simulation into routine practice, foundries can achieve higher consistency and lower costs, reinforcing the role of nodular cast iron in modern engineering. We encourage broader adoption of these methods to foster innovation in casting technology for nodular cast iron and other alloys.