1.1 Introduction
Metal spinning process is a metal forming process used in the manufacture of axisymmetric hollow products. Essentially, all spinning techniques involve rotating a workpiece clamped onto a chuck while the spinning tools approach the workpiece and deform it to the require shape[1]. The spinning process makes it easy to control the dimensions of products. The strength of products also increases during spinning process. Other advantages include a high material usage rate, fewer production stages, a lower forming force, and flexibility in manufacturing. For these reasons, metal spinning process has been widely used in various applications.
The term metal spinning refers to a group of three processes: conventional spinning, shear spinning and tube spinning. A common feature of the three processes is that they allow production of hollow and rotationally symmetric parts. The main difference between the three is apparent in the wall thickness of the workpiece [2]. In conventional spinning, a sheet blank is formed into a desired shape according to the contour of mandrel, and the wall thickness remains constant throughout the process (as shown in Figure 1.1). In contrast, in shear spinning the wall thickness is reduced while the diameter of the part remains constant. A blank of initial thickness t0 is reduced to a thickness t where the final thickness t is related to the wall angle α by the well known sine law (as shown in Figure 1.2) [2]. In tube spinning, also known as flow forming, a tube is mounted over a rotating mandrel. The roller pressing against the tube advances in the axial direction as the tube rotates under the roller. The wall thickness decreases locally under the pressure of the roller while the roller gradually advances through the tube surface (Figure 1.3). During tube spinning, the thinning of the wall results in elongation of the tube in the axial direction with no change in the nominal diameter [3].
2
Figure 1.1 Schema of conventional spinning
Figure 1.2 Schema of shear spinning
Figure 1.3 Schema of tube spinning
3
The above-mentioned classification for general spinning processes is widely accepted in literatures; however, in recent years new spinning techniques have appeared. These are not easily classified into conventional spinning, shear spinning and tube spinning, such as neck-spinning, splitting spinning, mandrel-free spinning, asymmetric spinning, etc.
1.2 Neck-spinning Process
Neck-spinning is a kind of spinning process used to reduce the diameter of cylindrical tube ends (as shown in Figure 1.4). In neck-spinning process, rollers displace to form the shape of tube ends. It is usually performed in multiple steps with symmetric rollers. For most applications, mandrel is not necessary during neck-spinning process; therefore, the shape and the wall thickness are controlled by the roller forming path. The mechanics of neck-spinning is quite different from conventional spinning, shear spinning, and tube spinning.
Figure 1.4 Schema of neck-spinning
In this study, the neck-spinning process was applied to form the neck part of high pressure tube ends. High pressure tubes have been widely used in various applications such as gas generants in airbag inflators, motorcycle airbag jackets, compressed gas dusters, soda
4
siphons, and cream whippers. Based on application purposes, various types of gas, such as CO2, N2O, N2, and Ar, are filled in the tubes. This study used a spun tube as a high pressure CO2 vessel, which is a component of motorcycle airbag jackets.
For this application of neck-spinning, the process should be performed at elevated temperatures. In high pressure vessel, the tube end is formed into domed shape and boss (as shown in Figure 1.5). The reduction of diameter is quite large at the tube end, so performing the neck-spinning process at room will cause a fracture to occur (as shown in Figure 1.6).
Therefore, in this study, the neck-spinning of tube should be performed at elevated temperatures to prevent fracture.
Figure 1.5 Schema of high pressure vessel
Figure 1.6 Facture after spinning at room temperature
5
1.3 Literature Review
Several researchers have conducted experimental and theoretical investigations on the influence of the various parameters on the spinning process [4]-[11]. Progress in computation capability and software coding has enabled the application of finite element analysis to the tube spinning process. Hauk et al. [12] used an axisymmetric model and a one-thirty sixth 3D model to simulate the flow-splitting process. Only a few steps of flow-splitting were simulated successfully using three-dimensional model due to large computation time and difficulties during manual remeshing; therefore, three-dimensional simulation of flow-splitting process at that time can hardly be applied to determine the proper process parameters. Iguchi et al. [13]
used a dynamic-explicit code DYNA-3D to analyze the spinning manufacturing process for exhaust system components of motor vehicles. The results showed the distribution of stress and strain which evolved in the material during spinning. This provided useful information for the prediction of failures during spinning. During spinning, the material temperature increased to as high as 300 ℃; however, changes in temperature and material properties were not considered in their simulation due to the computation cost. Hua et al. [14] used ANSYS to establish a three-dimensional elastic-plastic finite element model for the three-roller backward spinning of a cylindrical workpiece. The simulation results showed a variety of phenomena that occur during spinning. These included bell-mouth distortion, build-up, bulging in the front of and between the rollers, and diametric reduction and growth. Although both experiment and simulation of backward spinning were performed in their paper, the simulated results were not quantitatively verified by experimental data. Xia et al. [15] used MARC to simulate the process of multi-pass offset tube neck-spinning. Their results showed that the distribution of strain and stress was non-axisymmetric; the equivalent stress distributes and varies along the axial direction section by section and reached a maximum at the opening end of the spun workpiece.
The thickness reduction at the opening end, and the ellipticity and axial elongation of the spun workpiece increased with increasing spinning passes. The linearity of forward path spinning
6
was significantly less than that of backward path spinning. Similarly, the simulated results were not quantitatively verified by experimental data.
In the above literature, the tube spinning processes were all performed at room temperature. Figure 1.6 shows that performing the spinning process at room temperature will cause a fracture at the top of the tubes, especially when the deformation is large. However, few studies have mentioned tube spinning at elevated temperatures. Makoto et al. [16]
invented a new CNC spinning machine comprised of rollers with heaters. The heated rollers heated the magnesium tubes and formed them into various shapes by spinning. The forming possibility of magnesium tubes was experimentally demonstrated. However, the heated rollers are not suitable to apply to form the material at high temperatures because the rigidity of rollers decreases as the temperature increases. Mori et al. [17] developed a hot shear spinning process of cast aluminum alloy parts to eliminate casting defects and obtain a desired distribution of wall thickness. Hot air heated the blank during the shear spinning process to maintain the forming temperature at 400 ℃. The commercial software LS-DYNA was adopted to simulate the hot shear spinning but only the distribution of equivalent plastic strain was presented and the simulated results did not compare to the experimental data. Yang et al.
[18] established a 3D coupled thermo-mechanical FE model of the hot splitting spinning process of magnesium alloy AZ31. The influence of different initial temperatures of the disk blank and different feed rates of the splitting rollers on forming quality of deformed flanges was investigated numerically. However, no experimental data were proposed to verify the simulation results. In summary, finite element analysis has been successfully applied to the tube spinning processes, but no temperature effects have been considered on tube neck-spinning.
7
1.4 Objective of Present Study
Despite the above-mentioned efforts on introducing FEA into metal spinning process, a complete and accurate finite element model for neck-spinning process at elevated temperatures has not yet been proposed. Therefore, the objective of this study is to construct a comprehensive finite element model to investigate the tube neck-spinning process at elevated temperatures. Comparing the results of the simulation and the experiment would verify the finite element model. Furthermore, the verified finite element model will be used to discuss the influence of the roller feeding pitch and to investigate numerically the roller forming paths to improve the thickness distribution.
1.5 Research Method
In current industrial practice, production of high pressure vessel requires two processes.
First, the closed-bottom cylindrical tube is manufactured from sheet steel using multi-stage deep drawing process. Diameter of the tube end is then reduced using neck-spinning process (as shown in Figure 1.7).
Figure 1.7 Production of high pressure vessel
The deep drawing process is highly efficient for manufacturing closed-bottom tubes, but the tubes manufactured using deep drawing are unsuitable for this study. In finite element analysis on neck-spinning process, accurate material properties of the tubes are important.
8
However, it is hard to cut specimens from the tube due to its small diameter so that specimens can only be cut from original sheet steel. Hence, the work hardening effect of the deep drawing process will be ignored in the simulation. Moreover, only tensile tests can be conducted using sheet specimens but the internal stress states generated in the material during neck-spinning process are different from those during tensile test. To resolve the above problem, original material was changed from sheet steel to rod steel; therefore, the tubes for neck-spinning and specimens are identically manufactured from rod steel using turning and boring without other work processes, thus compression test can be conducted using rod specimen.
In order to construct an accurate and comprehensive finite element model for neck-spinning process at elevated temperatures, this study firstly performed material tests to obtain properties of the tube. Low carbon rod steel AISI 1020 was used in this study. Uniaxial compression tests were conducted at various temperatures and strain rates since the material is sensitive to strain rates at high temperatures. Neck-spinning experiments were then performed and the finite element analysis with the same process variables was also conducted by commercial finite element software, Abaqus/Explicit, incorporating these obtained material properties. After verifying the consistency of simulated and experimental results, a comprehensive finite element model for neck-spinning process of tubes at elevated temperatures was assured. Finally, the finite element model was used to find the proper process variables and to analyze the influence of roller forming paths on neck-spinning process.
1.6 Structure of Dissertation
This chapter introduces the background and application of the tube neck-spinning process at elevated temperatures. Chapter 2 presents the experiments of material properties and corresponding constitutive models of the flow behavior. Chapter 3 specifies experiments
9
of neck-spinning process. A finite element analysis of neck-spinning process and comparison between experimental and simulated results are discussed in chapter 4. The influences of process variables and roller forming paths are discussed in chapter 5. Finally, chapter 6 concludes and summaries this study.
10