Research on key casting process of high-grade CNC machine tool bed nodular cast iron

1 Introduction

As is well known, materials play a crucial role in the performance of machine tools [1,2,3]. The composition, production process, heat treatment process, and processing technology of materials will all have an impact on their mechanical properties [4,5,6]. Machine tools are often referred to as the “mother machines of industry.” High-grade CNC machine tools require a high level of dimensional precision control. Compared to the use of gray iron in the casting of bed bodies for large machine tools in China, foreign counterparts frequently utilize nodular cast iron (NCI) such as QT600-3 for producing high-precision machine tool beds. This choice is based on NCI’s ability to control the elastic modulus of the castings, thereby imparting higher rigidity and resistance to deformation. Consequently, NCI beds outperform gray cast iron counterparts in terms of precision retention and noise reduction [7].

Another crucial factor affecting the precision of high-grade machine tools is the residual stress generated during the casting process [5,8–10]. Large machine tool beds are often characterized by complex structures and larger dimensions. As a result, the casting process plays a crucial role in the quality of the bed structure and the success of the casting, and different casting methods also directly affect the distribution of residual stresses within the castings. Therefore, aside from post-aging treatments, optimizing the casting process and analyzing the internal stresses resulting from temperature field distribution are of paramount importance for reducing residual stresses. In practical manufacturing, mass production often necessitates trial casting to validate the selected casting scheme, which is not only inefficient but also incurs high costs.

With the maturation of casting simulation technology [11–13], visual simulations of the casting process are now achievable, allowing for an intuitive assessment of the merits and drawbacks of various process schemes. Finite element analysis of casting processes has been widely applied in scientific research and practical production, with its greatest advantage being the improvement of production efficiency and cost savings. Currently, the mainstream simulation software includes AnyCasting and ProCAST. Compared to ProCAST, AnyCasting has lower requirements for grid quality and can simulate complex castings. This software can simulate the filling and solidification processes during the casting process well, playing a good auxiliary role in adjusting and optimizing the process. Si et al. [11] combined the advantages of finite difference method and finite element method in the analysis of molten iron flow and stress analysis during the solidification process using specific algorithms to better simulate the casting process. Kwon and Kwon [12] used AnyCasting to simulate and optimize the gates and flow channels of automotive parts, predicting and reducing internal pore defects. Sun et al. [13] also used AnyCasting to simulate and optimize the casting of boom brackets for high-speed multifunctional rescue engineering vehicles. This enables real-time, targeted improvements to the adopted schemes. Hence, this study conducts a simulation analysis of the casting process for QT600-3 machine tool bed castings commissioned by a company. In addition, it investigates the microstructure and precision retention performance of the machine bed.

2 Materials and methods

In this article, the gravity sand mold casting process is used to complete the casting of a large integrated lathe bed for car grinding. The material used for the molding is resin sand, which has good flowability and is easy to compact. The sand core is made of alkaline phenolic resin self-hardening sand, which shapes the internal control cavity and parts of the casting that cannot be directly cast into shape. The core bone is embedded in the sand core to improve its strength and stiffness. The core bone material is round steel and cast iron. The mold used in this process is a wooden mold. The casting material is ductile iron QT600-3, and the molten iron is poured after smelting, spheroidizing, and inoculation processes.

A commercial software (AnyCasting) is used to simulate mold filling and solidification in casting and assist in optimizing casting processes. The process parameters are all from the production site, and the pouring system and cold iron settings have been improved based on the consistently used process parameters. By simulating and studying the internal defects of castings for improving the gating system and adding conformal cooling iron during the casting process. The suitably properties and microstructure of the obtained castings were observed through metallographic observation, tensile testing, and hardness testing.

3 Casting process design for the machine tool bed

The three-dimensional (3D) model of the machine tool bed from various angles is depicted in Figure 1. The overall dimensions of the bed were 4,580 mm × 1,659 mm × 744 mm, the lead rails were 4,515 mm long, and each individual rail was 111.5 mm wide. The grinding machine bed was oriented perpendicular to the lead rails and was offset to the left, creating a “品” (Chinese character “pin”)-shaped configuration. Internally, it consisted of ribs, cross beams, and cavities. The material chosen for casting was QT600-3, and the sand mold selected exhibited good stiffness and appropriate deformability, capable of withstanding the impact of molten iron while accommodating graphite expansion [14–16]. The existing pouring system makes the filling of molten iron more reasonable by reasonably calculating the head of the sprue on the existing basis. The position of the symmetrical sprue on both sides in the original pouring system has been changed to the ends of the guide rail, resulting in a more uniform filling of all parts of the casting. At the same time, according to the distribution of the lathe bed structure, the inner runner is changed from the original one ceramic tube to two smaller ceramic tubes, and two guide rails are poured simultaneously. The pouring rate of molten iron is 110 kg·s⁻¹, the tapping temperature of molten iron is 1,420°C, and the pouring temperature is 1,360°C.

Figure 1

3D solid modeling of the machine tool bed at various angles: (a) bottom view; (b) front view; and (c) rear view.

Due to the complex structure of the bed and the wedge shape of its bottom, a bottom pouring system was adopted for uniform filling. Design the pouring system according to the structure of the lathe bed. Based on the characteristics of the bed, it can be seen that the middle thick part bears the support function of the hanging guide rails at both ends. Therefore, the design of the pouring and riser should consider the size of the lift and whether it will cause impact damage to the mold. The horizontal runner needs to have the functions of stabilizing flow, distributing molten iron, and blocking slag, so a square tube with a larger runner area is used to feed molten iron at both ends. The inner runner uses cylindrical porcelain tubes to guide the molten iron to the guide rail position, adjust the temperature distribution and solidification sequence of the casting, and also assist in blocking oxide slag inclusion in the transverse runner. To prevent uneven cooling of the thick part, risers are uniformly set at these positions to accelerate cooling and avoid the problem of air entrainment during the filling process of molten iron.

Symmetrical pouring was performed at both ends using two ø90-mm straight sprues, and no sprue cup was designed since it was not used in actual production. The channels were eleven ø35-mm ceramic pipes for delivering the molten iron. Their inlets were evenly distributed in the vertical direction along the lead rails and the grinding machine bed section. Specifically, two sets of 2 × 4 ceramic pipes were distributed along two lead rails, while the grinding machine bed had three ceramic pipes arranged in the middle. This configuration ensured a slow and uniform filling of the mold cavity as molten iron flowed from both ends toward the center. The pouring system design is illustrated in Figure 2. The thicker sections of the bed casting are susceptible to shrinkage and porosity, and the most effective means of addressing this issue is to use risers for feeding during process design. During the pouring process, molten iron often carries oxide inclusions at its front end. Therefore, strategically placed risers can also serve the purpose of removing impurities and preventing entrapped gases [17,18]. In view of this, 2 × 9 risers were evenly distributed for venting and feeding in both the bed and the operating platform pedestal, ensuring a dense structure in the castings and preventing defects. Lead rails are a critical casting component of machine tool beds. The pouring temperature during production was set at 1,350°C. To ensure the timely conduction of accumulated heat on the lead rail surface and to reduce internal stresses caused by temperature differences, chills were spaced at 15 mm intervals near the ingates along the lead rails. These chills served to mitigate stresses and enhance the strength and stiffness of the lead rails, and their on-site configuration is depicted in Figure 3.

Figure 2

Design models of the gating system and chills: (a) gating system and castings; (b) pouring system (at both ends); and (c) castings and chills.

Figure 3

On-site layout of the pouring system and chills. The yellow pipes are the channels for pouring molten iron, and these channels can ensure uniform pouring to avoid erosion and pressure differences. The black blocks are cold irons for uniform cooling function.

4 Simulation of filling and solidification in casting process

4.3 Analysis of simulation results and casting process evaluation

Upon completing the calculations, the results were opened in anyPOST to visualize the filling and solidification processes. Figure 4 illustrates the temperature distribution at different locations of the castings during the pouring and molten iron filling process. The pouring gates were symmetrically distributed on the two sides of the lead rail centerline, but the right-side gates were closer to the ingates and had a shorter travel distance. Consequently, molten iron began to fill from the bottom through the ceramic pipe earlier on the right side. As depicted in Figure 4(a), bottom filling occurred at the right ingates when the overall filling reached 1.72%. From the temperature distribution in the two end runners in Figure 4(b), it is evident that the temperature of the molten iron at the front end of the ingates near the pouring gates was higher than that farther away from the pouring gates at each end. This is because the molten iron in the ingates near the pouring gates had a shorter travel distance and received flow replenishment more quickly. When the filling volume fraction reached 18%, the molten iron from both left and right ingates converged, initiating the integrated filling process, as shown in Figure 4(c). Prior to this, owing to the relatively slow filling rates of the two separate sections, the solidification rate was comparatively faster. However, after the convergence of the two sections, the filling rate exceeded the solidification rate. Therefore, the filling volume fraction of 18% marked the watershed between filling and solidification. The filling was completed in 90.943 s, while complete solidification required 787.62 s. When the filling reached 100%, the solidification fraction was 27.07%. As observed in Figure 4(d), the bottom lead rail section, which was poured earlier, had already solidified to a significant extent. This is attributed to the rapid cooling effect of the chills, which ensured timely and uniform solidification of the lead rail surface.

Figure 4

Temperature field distribution of the castings under different filling conditions: (a) 1.72%; (b) 2.63%; (c) 18%; and (d) 100%.

The current casting design exhibited a smooth and adequate filling process with a certain feeding effect. However, further optimization of the gating system was required to reduce the solidification rate in the thicker section of the bed base. After reaching a solidification fraction of 51.85%, some risers had almost completely solidified, leading to a decrease in their feeding function. At this point, the temperature in the thicker section of the bed base was below 1169.514°C, which is lower than the liquidus temperature, as shown in Figure 5(a). Residual stresses in the bed arise from multiple sources, among which the primary contributors related to temperature during the casting process are internal stresses and phase transition stresses caused by significant temperature gradients during the solidification process [23]. The temperature field distribution during filling and solidification reveals that, after filling was complete, the overall temperature gradient in the casting was not substantial. The maximum temperature gradient occurred within the runners, and the risers also solidified sequentially as desired. Therefore, the primary concern was the temperature differential within the bed. In the later stages of casting, the overall temperature differential within the bed remained relatively small. The simulation results indicate that after filling, the temperature differential stayed within 20°C and gradually decreased as the solidification process proceeded. Moreover, the lead rail surface solidified uniformly, with no significant temperature gradients in the horizontal plane, which is beneficial for reducing internal stresses on critical machining surfaces. As we know, a strong temperature gradient will generate residual stress. Due to the different temperature gradients during the solidification of molten iron, the solidification or crystallization will be different. This transformation from liquid phase to solid phase will result in volume shrinkage changes, resulting in a large amount of residual internal stress [24].

Figure 5

Temperature distribution of the castings at solidification fraction of (a) 51.85% and (b) 95.98%; prediction of shrinkage defects on (c) the base and (d) the lead rail surface.

Shrinkage and porosity are additional key indicators for evaluating the casting process. The feeding effect of the risers in the later stages of filling and solidification was limited. The wedge-shaped risers were chosen based on practical casting experience, as this shape facilitates mold removal and rapid cooling, especially in the fine neck sections. This allows for self-compensatory shrinkage through graphitic expansion in the later stages of solidification. As observed in Figure 5(a) and (b), the temperature distribution within the casting was uniform in the later stages of solidification, with slower cooling in the thin-walled and protruding areas. Since the lead rail surface was equipped with chills, no risers were used for compensatory shrinkage, as evident from Figure 5(c) and (d), where no shrinkage or porosity defects were present on the lead rail surface. However, sporadic areas on the base showed tendencies for shrinkage and porosity. These areas can often be optimized through subsequent improvements to the riser design, and in practical production processes, graphite-induced self-feeding can also be achieved by adjusting the composition of the molten iron. In addition to predicting and intervening in casting defects through simulation methods, it is also important to combine modern detection methods for surface and internal defect detection (i.e., nondestructive testing) of actual produced products, which is an important part of quality functional control. Common nondestructive testing techniques include ultrasonic testing, magnetic particle testing, electromagnetic testing, etc., which can effectively detect the location and distribution of defects.

5 Casting microstructure analysis

This casting adopts sand casting process, and the parting surface is determined based on the pouring position. Sand cores are used to fill the internal cavities, holes, and parts of the casting that cannot be directly cast into shape. Due to the complexity of the inner cavity, a core assembly design is adopted. Embed the core bone in the sand core to improve its strength and stiffness. To ensure the exhaust of the sand core, an exhaust channel is set up in the sand core during core making. The wood mold in this process is relatively complex, so it is manufactured using wood molds that are easy to process and low in price. Figure 6 presents the finished product image of the castings after shakeout. In Figure 6(a), it can be observed that the pouring system and risers were connected to the castings after pouring, and irregularly solidified thin sheet-like structures remained at the parting line. The physical appearance of the bed after removing the gating system and other residual structures is shown in Figure 6(b). Upon visual inspection of the casting exteriors, no defects such as cavities, missing features, or incomplete filling that could lead to casting failures were evident. The microstructural details were well-formed, indicating good structural integrity. To mitigate residual stresses during the casting process, the bed casting would undergo aging heat treatment after the removal of the casting-associated structures. In addition, a solidification process employing 3D vibration was utilized during casting [25] to minimize residual stresses and enhance the mechanical properties of the large ductile iron bed base. The use of 3D vibration during the solidification process is an advanced method to reduce residual stresses after the whole liquid phase was transited to solid phase. The high-frequency vibration is beneficial for eliminating small strains between grain boundaries, thereby reducing the residual stress of the entire casting after complete cooling [25].

Figure 6

Site view of the castings (a) before and (b) after the removal of the gating system.

For a closer examination of the casting microstructure, standard Y-shaped test specimens were cast simultaneously in the same batch with the bed casting. Mechanical testing and microstructure observation were conducted on samples taken from the cast Y-shaped test specimens. After grinding and polishing the obtained samples, the polished surfaces were immersed in a 4% alcoholic nitric acid solution to reveal grain boundaries. Subsequently, the polished surfaces were observed under an inverted metallurgical microscope. The microstructures are shown in Figure 7. Scanning electron microscope (SEM) images in Figure 7(a)–(c) reveal well-dispersed spherical graphite within the microstructure, with good sphericity and a spheroidization grade of 2–3. Statistical analysis indicates a spheroidization rate exceeding 85%. Fine graphite nodules were uniformly distributed on the lamellar pearlite matrix, and no porosity defects were found in the matrix. The matrix structure was dense, and there was a close bond between the graphite nodules and the matrix, with no signs of detachment or gaps between them. Optical microscope images in Figure 7(d)–(f) show that the matrix structure mainly consisted of lamellar pearlite, and the content of phosphorus eutectic + cementite was ≤0.5%. The interlamellar spacing was extremely small, thereby enhancing the atomic bonding between ferrite lamellae and carbide lamellae and increasing the elastic modulus of the ductile iron casting [26,27]. The elastic modulus of the ductile iron bed, however, was closely related to machine tool precision and low stress tolerance.

Figure 7

Microstructure diagrams: (a)–(c) show SEM images at different magnifications, and (d)–(f) show metallographic microscope images at different magnifications. The spheroidization rate of graphite exceeded 85%, with a spheroidization grade of 2 to 3, and the matrix structure consisted primarily of lamellar pearlite.

The compact structure of a large integrated lathe for car grinding makes the mold structure, sand core shape, and pouring system more complex, resulting in defects such as shrinkage, porosity, and air entrainment in some positions. AnyCasting software was used to conduct numerical simulation analysis on the filling and solidification process of lathe bed castings, predicting possible situations such as casting air entrainment and isolated liquid. Then, the position of the riser and the shape of the cold iron are reasonably designed, and the cooling and solidification sequence of the ductile iron is adjusted. At the same time, it also plays a role in eliminating isolated liquid phases. The mold is divided into upper and lower parts for mold opening, and the lower mold is distributed with a pouring system and a casting cavity. This open pouring system is conducive to exhaust and the discharge of oxide inclusions. Set a riser for exhaust at the bottom to avoid setting a riser on the machining surface.

In order to verify the simulation results, we optimized the original casting process in the factory production based on simulation and reprocessed the optimized products through trial casting. The testing parts from the trial castproduct indicated that it is easy to form gas entrapment and internal stress due to large thermal gradients. Figure 8 shows the tensile test results and Vickers hardness tester test results obtained by sampling the casting guide rail. In order to compare the advantages of this technique, we selected the same material from conventional casting as the control group, as shown by the C1–C3 curves in the figure. The optimized sample is shown as the QC1–QC3 curves. In each group of tests, we set three parallel samples for error assessment. The tensile mechanical performance of conventional casting specimens indicated that the tensile strength is 405 ± 22 MPa, and the elongation is about 3.5 ± 0.2%. However, the tensile strength is 468 ± 28 MPa and the elongation is about 3.6 ± 0.3% obtained from our proposed method. The maximum tensile strength of commercial ductile iron prepared by conventional casting is about 440 MPa, and the minimum tensile strength is about 412 MPa. Compared with the performance of commercial ductile iron, the casting strength after optimizing the casting process is much higher. From the tensile part of the casting, it can be observed that the matrix cleavage characteristics exhibit strong toughness characteristics, while there are also many ductile dimples, and graphite is uniformly distributed in the cast iron structure. After polishing the back of the stretched sample of the casting, its hardness was tested using a Vickers hardness tester. The average of three tests showed that the hardness was 220 HV, meeting the performance requirements of the lathe casting.

Figure 8

Mechanical performance testing of lathe bed castings: (a) tensile fracture samples; (b) sample fracture; (c) tensile time stress curve; (d) and (e) fracture morphology; and (f) Vickers hardness test.

6 Conclusions

The simulation analysis is conducted on the casting process by using AnyCasting for large-scale machine tool bed casting. So it is possible to prepare good casting process parameters to the molten iron filling pattern of the gating system, which provides guidance for the actual production process. The finding indicates that this process exhibits uniform and stable filling capabilities. The results of the filling and solidification simulation demonstrate that the inclusion of chills in the machine tool lead rails and grinding machine bed facilitates timely heat transfer, reducing the heat accumulation from the bottom-pouring casting process on both components. Consequently, the thicker sections of the bed casting and the lead rails can solidify more quickly. After filling, the solidification rate reaches 27.07%. At this point, the temperature distribution within the casting is more uniform. In addition, the entire casting is below the liquidus temperature, effectively minimizing the internal residual stresses caused by significant temperature gradients. The microstructure of the castings consists of fine lamellar pearlite matrix, which contributes to favorable mechanical properties and dimensional accuracy retention for ductile iron castings. The internal structure of the castings shows a dense matrix with no shrinkage or porosity defects. The spherical graphite is uniformly distributed within the matrix. This structure may effectively resist deformation and make this material suitable for high-strength, low-stress bed castings.