Loading of Watts Link Frame Attachment Brackets 1

[Joest]

During the course of the last few months I have been designing a rear suspension system employing a solid rear axle, 3-link torque-arm suspension, and a centered watts link. The watts link has an adjustable pivot height for various roll center locations. The suspension will be sprung with coil-over dampers in the conventional fashion.

The primary function of the vehicle will be open-track road racing. I do, however, plan to drive the car on the street occasionally for recreation.

One of the primary concerns I've had is how much factor-of-safety I should build into the brackets that attach the watts linkages to the frame. It is straight forward to assume that the vehicle will corner at 1G and work from there, but I am more interested in what happens when I lose it and take a trip through the gravel trap or something similar. Does anyone have any sound advice they have learn from experience or have picked up working in the automotive industry as an analyst? Much of my work has been done with solid 3D models of the design (ProE) and I have also gone as far as analyze the brackets with several FEA/FEM packages such as Pro/Mechanica and Design Space.

If you need clarification of the layout of my design then let me know and I will send you a jpeg image.

Thanks in advance.

 

[GregLocock]

1g is a good place to start, but as you surmise it will probably be exceeded for example when you curb the wheel in a spin.

The good news is that a 'normal' factor of safety will give you a sufficiently robust design.

The reason is that most of the impact is absorbed by the tyre, and the 'softened' load gets split between both rods.

So if you design each rod and bracket for 1 g you /should/ be safe, with about a 50% safety margin.

This result was froma a driving test. Hitting the wheel sideways with a pendulum of axle loading mass at the same lateral velocity gave a completely different result, but inspection of the film we took revealed it was a completely stupid (unrepresentative) test. It was fun to try.

Cheers

Greg Locock