I've got a copy of SusProg...

and I'm not afraid to use it. Reading through the regulations of the RGB championship, there's no ability to tune the engine from when it got separated from the motorcycle. So if you're going to build a successful racer, aside from a large amount of driving talent, you need to find other ways of gaining a competitive advantage over your fellow competitors. 'Proper' aerodynamics aren't allowed (and although there doesn't appear to be a limit on underbody aerodynamics, the minimum ride height of 75mm seems to not be of much help in generating some venturi-assisted downforce), there are minimum weights (although getting close to them seems to be a challenge for those who do indeed take the RGB regulations at their word rather than constructing a specialist lightweight chassis) and everyone is running the same tyres.

So it appears the only way to beat your fellow racers is to out-drive them, or for those with less talent, out-handle them. Here is where SusProg comes in. If you're incredibly rich, you can spend the GDP of a small African protectorate on kinematic simulation software like MSC Adams to help with your suspension design so that you don't have to rely on Colin Chapman's ethos of 'the suspension works great if you don't let it' to avoid all sorts of nasty kinematic and alignment problems.

As a former ride and handling engineer (OK, so most of the vehicles I helped engineer were green, extremely heavy and capable of dealing death over large distances), I'm afraid I treat the rest of a car as a black box - if it doesn't work, you get rid and find something else that does. I'm going to design the car from the wheels inward - create a suspension design that works, then stick a chassis on it. Using SusProg is key to this. It lets you design the suspension virtually and see how it works before you commit yourself to cutting your first bit of tubing.

I've already made a conscious decision to avoid using someone else's uprights to hang the wheels off. For a start, I haven't got the time or inclination to find some suitable uprights, sort out suitable brake discs, callipers and wheels so that they all work together and, more importantly, fit and then design a suspension that fits in with the limitations of the upright. I plan to design easily (at least easily if you've got access to a nice CNC milling machine) build-able (possibly even fabricate-able) uprights and then hang everything else from there.

Enough with the planning, lets get on with the design. I'll start at the back, not least because I only have to deal with the wheel moving up and down and not with having to steer as well. Using a double wishbone design is the obvious choice (everyone else designing racing cars seems to do the same thing) but then it all gets a bit more complex. We need to control what happens to the wheel when it rises up and down relative to the body. The two things we need to control are camber (the inclination of the wheel relative to the vertical) and toe (the rotation of the wheel about the vertical).

Let's start with toe. If your wheel steers itself as it goes over bumps this may not be a good thing. If there is enough self-steer the car will potentially be less stable. A less stable car leads to a less confident driver and a less confident driver is a slower driver. Sometimes this steering over bumps can be a good thing - Lotus use it on the Elise to improve stability with the steering that results counteracting the grip changes due to the change in loading on the contact patch. If you can find a copy of IMechE paper C466/014/93 (it's in the proceedings of the Vehicle Ride and Handling conference - take a look at C466/010/93 as well if you get the time as it's a report on what I was doing at the same time) it's all 'explained'. Personally, I think eliminating toe change altogether is an easier objective and hopefully a racer will prefer an absence of any taint to the feedback he gets from the car.

Camber is a bit more subtle. If you tilt a wheel over, instead of a nice oval contact patch between the tyre and the road, you get one shaped a bit like a banana. As the rubber rolls through the contact patch it is forced in a curved path rather than a straight one and this deviation provides a small sideways force. This is handy if the force acts in the direction you want to turn and less handy if it doesn't. As the suspension moves, it moves through some form of arc and this changes the inclination of the wheel and can lead to reductions or increases in the grip of the tyre. The arc is subtly different between the wheel moving in bump (where the chassis stays flat to the ground) and in roll (when the chassis tilts moving the suspension mounting positions relative to the ground)

Another thing to consider is traction. Potentially we are going to feed 180 bhp to the rear wheels and we want as large a contact patch as possible to avoid the wheels spinning and wasting acceleration potential. A cambered tyre has a smaller contact patch than one running upright on the road. The first racer on the gas in a corner carries the most speed out and has the lower lap time, so being able to put the power down with the car rolled over is a key issue to resolve.

So taking all this stream of consciousness out, I've come up with the following design parameters:

• Minimal camber change in roll (I'm hoping for zero change if possible, like the Porsche Carrera GT)
• A non-migrating roll centre (I'll talk about roll centres in a later post, but they're very important to vehicle handling)
• Low camber change in bump, and a negative camber change if possible
• Zero change in toe with bump

OK, so we've got some parameters to work with, let's start messing with SusProg...