If the inertia of the wheel assembly is high, it will take more time for the sys tem to recover

from a disturbance such as a bump on the track, and thereby not allowing driver to exploit the performance from the vehicle. Therefore for any components in the wheel assembly, weight carries extra significance. Aside The upright has to incorporate all the pivots needed by the suspension and allows the system to move within its designed range of travel wit hout obstructions and cause binding. Other physical limitation includes the choice of the bearing sizes and the amount of suspension adjustments needed to be built-in to the design. The goals are being able to package the necessary components at their correct location and orientation within the confine of the w heel. At the same time, the types of construction, material, sizing of bearing and piv ots, as well as adjustment method will have to be decided. The spherical bearing offers tolerance for angular misalignment, which is required for the suspension system throughout the suspension travel ran ge, as well as allowing rotation in its axial direction. Sizing of the bearings is primarily based on the manufacturer s documentations against the known loading condition. the bottom ball joint becomes the primary source of loading in the upright. This allows for a reduction in spherical bearing size for the upper ball joint. The upper ball joints are loaded in double shear to ensure the reliability of the aluminum clev ice, while the bottom ball joints are loaded in single shear to provide for extra cle arance for the steering requirement. The fixture itself also needs to be of substantial structural stiffness to ensur e zero distortion when the parts are bolt onto the fixture and undergoes post weld ing stress relief heat treatment. The fixture is manufactured using manual machining method on a 3-axis milling machine to ensure geometric accuracy of the fixture components, and they are assembled using dowel pins to facilitate accurate assembly of the fixture compon ents Critical geometry such as the bore for the wheel bearings are postmachined after the heat treatment process as to ensure the thermal distortion of the heat-treating process will not effect such critical dimension

Since real world loading conditions and constraint can be incredibly complex, simplified representative conditions are often used to model the real w orld constrain. Fake parts are usually including in the FEA model to replace those parts in which may exist on the real assembly but their performance are not important to the model of interest. The results of the FEA then require the designer to inter pret with that knowledge in mind. In Pro/Mechanica, the sheet metal faces on the upright are analyzed using shell element , which is used specifically to model parts with thin cross sections. Whil e certain internal features and bearing bosses are modeled as solid elements. Forces are applied 17 from the tire and transmitted through the wheel to the hub and applied on the up right through the wheel bearing mounts. While upright is being constrained at the uppe r and lower ball joint by the suspension links that connects to it. With multitude of connection joints and complex part geometry, it would not be feasible to analyze the accurate reproduction of this setup. Thus a simplified model will have to be use d instead. The following simplifications are used to in the FEA model to represent this type of loading condition. The actual upright uses deep groove ball bearing made by NTN bearing. The comple x interaction between the ball bearings and its multitude of contact region is not what the focus of the upright design. The representative bearing is made of a solid part of the exact dimension of the actual bearing. To compensate for the reduction of structural stiffness comp ared to a solid part as seen in the real bearing, the FEA bearing is modeled with a r educed Young s Modulus as opposed to that of the solid steel components Section 2.b.2) Hub, Wheel, and Tire: In the real world the cornering loads are transmitted by the combination of the 3 parts to the upright assembly. What upright itself sees is a force and a moment load. Therefore the complex relations between the 3 parts are not important for the re sult of the thesis. To replicate the loading condition, a simplified FEA part is made to represent the 3. The part consists of a fake hub portion that is connected to th e upright 18 via the FEA bearings, and a long moment arm that extends to the tire contact pat ch center location of the actual tire, at which point the loads are applied to the model

(Fig F2). This part is modeled as solid steel part to minimize its own deflectio ns and transmits all the loads to the upright assembly. Section 2.b.3) Welded Joints: Although the fabricated parts are consists of primarily welded joints, in the FE A the relationship can be hard to replicate. The assumption made is that since the upr ight will be stress relieved in post-welded state, the joint condition will be homoge nous to that of a regular material. Also, in using the Pro/Mechanica, the program has so me difficulty in dealing with joints that connects solid elements to shell elements , and with most welded joints being welded along multiple seams such a connection will be difficult to model. Therefore the joints in the model are kept as homogenous connection, and the results in those areas are not used in the analysis. Under load, the a-arm will resist the movement in lateral and longitudinal direction, while the pushrod will resist the load in the vertical d irection. For upper ball joint, since there are no pushrod connection, it resists movement only in longitudinal and lateral direction, therefore it is assigned wi th constraints in x and y axis. For the steering/toe-link pickup, the only link that connects to this joint is either the steering link or toe-link. They only resist movement in the lateral direction, so only y-axis is constrained in the model Since it is important to know the yield and material limit, as well as the compu tation of safety factor, Von Mises stress is used in presenting the stress results. The FEA results are compared against the fatigue strength of the material corrected for a known service life. The correction factors followed that of a standard fatigue calcula tion and takes into account of load factor, size factor, surface quality, operating tempe rature, and reliability. For full calculations please refer to Appendix E. Section 2.e) Optimization Parameters: The deflection of the upright assembly will be the basis for the optimization pr ocess. With stiffness being the performance standard and weight being the concern, the design goals are defined to be reduction in weight over the 2006 design with comparable stiffness. To optimize for weight, sheet metal thickness for differen t faces of the upright are changed iteratively based on the previous run s stress distribu tion and deflection value, the material thickness were reduced in the areas where str esses are low. The limiting factor being stresses cannot exceed the material limit. Wi th

available thickness value based on available stock material, a number of combina tions were analyzed and the optimum front and rear upright designs were selected as th e final designs. Section 3: Manufacturing The finished uprights are then subjected to stress relieving process to alleviate in duced thermal stresses in the welding process. The post-treatment uprights are then po stmachined to achieve the required dimension for the wheel-bearing bore. Section 4.a) Non-Destructive Testing: Dye-penetration testing is a common non-destructive method to validate the quali ty of weld. Dye-penetration test provides a simple visual method to allow for quick post welding inspection for quality of weld. This is based on the fact that the inside and outside wheel tracks curves of different radii. This is especially pronounced 2 Figure C3: Toe/Alignment Figure C2: Ackerman Steering Principle when the corner in question is tight, and that the distance between the inside and outside tires(also known as the Track) takes up a significant percentage of the radius of the curve. This is adjusted on the upright by different set of pickup point available on the clevice In bump this is governed by the relation between the upper and lower control arm and their respective length. In steering this will be governed by the caster and KPI. On the front this is set by adjusting the length of the steering linkage. And on the rear this is set by adjusting the length of the toe link. The amount of camber compensation is based on suspension kinematics and design.