In order to accurately anticipate the stress experienced by the wheel centers during spirited driving, telemetry data from previous years' competitions was analyzed to determine the maximum stresses experienced at the tire-road contact patch at each of the four wheels. 

 The force acting on the tire from the road surface was decomposed into its X, Y and Z components and the resulting stress from each component of the force was calculated to check for potential failure of the bolts, wheel center and nuts.

The goal was to torque down the bolt enough to maintain a high enough preload between the bolt and the wheel center such that the frictional force was sufficient at peak cornering loads to prevent the wheel center from slipping. 

 Equations for bolt preload, tightening torque, von mises stress etc were obtained from Shigley and this helpful website

FBD equations  were transferred to excel to quickly iterate on the dimensions of the spokes, nuts and bolts to arrive at ballpark figures that we could use when designing the wheel center in NX. This saved a lot of time in FEA.

The devil is in the detail! One should not forget to consider the +/-25% accuracy of the torque wrench.

Once Von Mises Stresses were obtained for the wheel centers, a Decision Matrix was created to compare the pros and cons of using Al, Mg, Fe and Ti. 

Goal: Obtain a material that balances cost, mass, toughness availability, corrosion resistance, yield strength (TYS), and Youngs Modulus (E).

Decision: Aluminum 7075-T651 due to low cost, low-ish mass, high toughness, corrosion resistance, acceptable yield strength (based on hand calcs) and high stiffness

Interesting find: 304 Stainless has a higher TYS than Grade 2 Ti

To avoid part clashes within assembly, space-fill CAD (keep out zones) were created. 

Wheel center space-fill CAD was created and wavelinked into the layout file. This allowed us to place each component in its rightful place within the layout file. Any subsequent modifications to a part (adding chamfers, holes etc) would only propagate downstream to the final assembly, preserving the layout file’s original design. 

This was a crucial design work flow in Siemens TeamCenter PLM that greatly improved the team’s productivity.

• First iteration of the wheel center based on hand calcs and excel results. At this point I was just making sure it passes the eye test.

•  This 1st version was rejected as it was deemed not manufacturable with our HAAS VF2 3 Axis CNC machine.

I simulated the X, Y and Z force on the tire-road contact patch and connected the contact patch forces to each of the 12 smaller bolt holes using RBE2 rigid spiders. 

Deflection across the wheel center's face in degrees was simulated. The target to hit was 0.5 degrees, or <25% of camber.

Von Mises stresses were calculated. A degree of skepticism coupled with engineering intuition is beneficial when analyzing FEA results. Therefore, convergence studies with incrementally decreasing mesh sizes were performed as a sanity check to make sure our FEA results did not change with tetrahedron size.

Modal analysis was ran to check for excessive vibrations due to resonance frequencies during racing conditions. 

Each nut was modeled in CAD to check for clearance between nut and ribbing. This allowed us to maximize the width of each rib to minimize deflection of each spoke.

We went through about 6 rounds of FEA. 6th iteration of the wheel center (left). Ribbing and fillets (in grey) have been added based on FEA results. Locating feature in the center of the wheel center. The 3x bolts/nuts are only to provide preload, never for locating.

Once we were satisfied with the FEA results, I proceeded to write g-code in NX CAM software. The challenge was selecting the correct tool, and setting correct feeds and speeds to prevent chatter/ tool breakage, and improve surface finish. FS Wizard was the go-to software to calculate those numbers. Also, since we had to make 16 of these wheel centers, I had to design the CAM so that tool changes were minimized, and feeds and speeds optimized for max efficiency. We used approximately 130 inch/min and 10,000 rpm for most tools.

The limitations of our HAAS VF2 3 Axis CNC machine meant we had to design and manufacture a jig to hold the wheel center in place when we flipped it over to mill the reverse face. A bolt was machined in house on the lathe to locate the wheel center precisely and clock it, without over-constraining it.

Workpiece set up in CNC. Toe clamps were used to secure a 3-jaw chuck which held the workpiece in place.

Finished wheel centers. Overall I was quite pleased with how they turned out. It took 3 attempts to manufacture it due to issues with surface finish/ chatter and bugs in the g code.

Wheel Centers after sandblasting and polishing and assembly onto the brake rotor. The locating feature in the center of the wheel was measured using a telescoping gauge while the wheel centered was still fixtured. The size of the locating feature was adjusted by altering tool wear offsets and using circular interpolation with a 0.5” endmill

Design and build a test fixture to simulate lateral and transverse loading on wheel center during hard cornering