2. FE analysis model for assembling wheel hub bearing assembly Fig. 1  shows the  schematic drawing of a wheel  hub bearing assembly of which manufacturing approach is studied in this paper. This assembly is fabricated by assembly of a hub machined after forged and a hub bearing unit by a cold rotary forging process. Fig. 2 shows the mechanism of the rotary forging process for assembling the hub and the hub bearing unit. During being assembled, the bearing unit can be fractured if the excessive  forming load is exerted. Also, the preload endowed in the hub bearing during assembly process has a strong influence on the performance of the assembly in its service because it can affect the wear characteristics of the assembly. Thus the forming load should be properly controlled.  In addition, because  the end of the hub is much elongated in the circumferential direction during the assembly process, it sometimes exposes to ductile fracture due to the cumulative  damage. For this reason, the contact condition between the end region of hub and the inner race of hub bearing should be optimally controlled and thus Unfortunately, this kind of rotary forging processes requires too much computational time to be simulated using a full domain analysis model because they have very long strokes compared to a conventional forging. It should be noted that the plastic deformation occurs due to quite small local contact of the forming die with material in the early stage, implying that the plastic deformation has a restricted effect on the neighborhood of the local contact region. This fact can be observed from Fig. 3, indicating that the contact region during rotary forging for assembling a wheel hub bearing assembly is quite small. Therefore, as shown in Fig. 4, a partial analysis model defined by two artificial planes of symmetry is proposed for the engineering analysis model in this paper, which was successfully applied for simulating a flow forming process (Cho et al., 2011). Of course, it is noted that the analysis results of this model may be more or less different from the real phenomena, especially at the planes of symmetry. However, it can be expected that quite reliable predictions can be obtained for the present rotary forging process because the plastic deformation occurring at the local contact area has little influence on that at the opposite side as shown in Fig. 3, emphasizing that the effective strain rate distribution is concentrated around local contact area. To check validity of the proposed approach, a rotary forging process of Fig. 4(a) was analyzed, which was previously studied using a hexahedral element (Moon et al., 2007). The shape of a preform and its dimensions are shown in Fig. 4(b).
The lower part of the preform is far away from being plastically deformed and its displacement was constrained by a constraint box in which all nodal degrees of freedom of the nodes are constrained. Flow stress of the material used is 0.135520.0(1.0 / 0.001) VH   
  MPa. The upper die revolves without any power exerted and the friction between the upper die and material was thus neglected. If the friction is considered, the revolving velocity should be an unknown, which may enhance a negligible solution accuracy at a great expanse of computational time. Fig. 4 is a 60° analysis model, which is composed of two artificial planes of symmetry, a part of material defined by them, a constraint box and tools or dies. First, to reveal the size effect of the analysis domain defined by two planes of symmetry, 30°, 60° and 90° analysis models were investigated.  Fig. 5 shows the predicted configurations of the deformed material together with the inner race of hub bearing unit at the selected planes for the 60° analysis model. As shown in the figure, there exists a non-negligible difference in deformed shape of material around the bent region between the mid-plane and symmetric plane. It can be seen that change of the shape is stationary round the mid-plane between 20° and 40° planes, implying that the mid-plane is quite far away from the effect of the assumed artificial plane, i.e., end-effect. Fig. 6 shows the history of deformation of sections A and B for the 60° analysis model. It can be seen at a glance that the deformation history of section A is nearly the same with that of section B as a whole. However, around the final stroke a distinct difference in contact region between the two planes can be observed. It is noteworthy that the size of cavity formed between the hub and inner race of hub bearing has a strong influence on the forming load when the process is controlled in terms of displacement (Shim et al., 2012) because the free surface around the major deforming region becomes very small at the final stroke.   Fig. 7(a)-(c) compares the deformed shapes at the mid-plane for the 30°, 60° and 90° analysis models, indicating that all the predictions are nearly the same. Comparison of the predictions in Fig. 7(a),  (b) and (c)  with the experiments in Fig. 7(d) shows that they are acceptable. Less than one hour of computational time was taken for the 60° analysis model. Fig. 8 shows the predictions obtained by the 60° analysis model at the final stroke with emphasis on finite element mesh system used. 3. Conclusions A computationally efficient finite element analysis model was proposed for analyzing a rotary forging process. 轮毂轴承有限元分析模型英文文献和中文翻译(2):http://www.youerw.com/fanyi/lunwen_33988.html