I am so insanely excited about the current project. This has been an idea that I have had for well over a decade and the time is finally right to pull the trigger and make it a reality. I am finally going to address the fundamental shortcomings of the G-body rear suspension and I’m going to share it with the world. There are plenty of products on the market to improve the rear suspension, by reducing deflection, improving anti-squat for launches, and reducing bind with spherical joints. But when it comes to cornering performance, there’s just no getting around the fact that the rear roll center is just way too high.
What is a Roll Center?
Briefly, a roll center is the point on the chassis where the sum of the forces from the tires in the Y-Z plane (see picture to right) are reacted and that the chassis wants to roll around. Imagine you welded a pin, aligned front-to-rear, on a post welded to the top of the rear axle housing. Then you added holes to the chassis to accept that pin. The car would roll around this point. And ride really rough! Now imagine driving along a straight path and you jerk the steering wheel to the right. As the rear tires generate lateral force, that force is transferred from the axle’s pin into the chassis, pulling the car into the turn. That lateral force at the pin, being well above the ground, creates a moment (torque) arm that transfers tire load from the inside to the outside tire. This happens immediately, before the body even begins to roll. The higher that pin (or roll center), the more load is instantly transferred.
If the roll center was the same height as the vehicle’s sprung weight’s center of mass, the suspension wouldn’t want to roll at all. (And if the line passing through the front and rear suspension roll centers passed through the center of mass, the suspension wouldn’t roll.) The effect of the roll center height in this case is similar to an anti-roll bar, in that it is creating unequal vertical forces in the left and right side of the suspension and resisting roll. Instead of that unequal vertical force going through sway bar end links, it is actually in the control arms. Conversely, if the roll center is at ground level, there is no roll resistance due to forces in the control arms, and all tire lateral load transfer (LLT) from the inside to the outside tire is due to body roll creating unequal spring forces, unequal damper forces, and twisting the anti-roll bar. The difference is it takes time for the body to roll, and therefore it takes time for LLT to build up with a low roll center.
Tire Lateral Load Transfer
So by now, you might be wondering why LLT matters so much. And the answer is because you don’t get a constant “return on investment” with tires. As you add load to a tire (vertical downward force, or weight), you don’t get a constant increase in lateral force capability. The more load you add, the worse your rate of return is. The result across an axle is that as one side unloads, it loses more lateral force capability than the other tire gains as it loads up. This is why a low center of gravity and wide track width are favorable for cornering performance; both reduce LLT.
This concept is critical for chassis tuning. By making the front suspension resist roll more than the rear, you can force the front tires to take a larger share of the weight transfer during cornering. This reduces total front grip and increases rear grip, making the car “tight” or more difficult to turn in. Alternatively, if the rear tires take on more of the LLT share, the car becomes “loose” or more responsive but potentially unstable. This concept is referred to as lateral load transfer distribution (LLTD), or often tire lateral load transfer distribution (TLLTD).
Getting Back on Track
Okay, so now let’s combine roll center, TLLT, and the G-body in particular. The G-body front independent suspension has a very low roll center. Somewhere in the neighborhood of 3″ (at least on my particular setup, others won’t vary too much). The rear suspension, on the other hand, is more like 18″. So imagine what happens when you turn into a corner hard. The front suspension has very little LLT initially as it waits for the body roll to create LLT. The rear suspension generates most of its LLT as soon as the rear tires generate lateral force. This is because the roll center height is very close to the vehicle’s sprung weight’s center of mass. So the result is the front end grips harder than the rear until the body roll sets in, but by that point it is too late and the rear end has started to slide. In practice, the driver learns this chassis behavior and limits how hard they turn in to prevent spinning out or loosing time trying to correct the slide. If the rear end behaved more like the front, you could turn in harder and the rear would stick.
You do want the rear roll center to be slightly higher than the front. The front tires generate lateral force first, so the rear should react faster to catch up. You just don’t want such a drastic difference as is seen on the G-body (or really any other triangulated 4-link paired with an independent front suspension.) So how do you reduce the rear roll center height?
Well I’m glad you asked. The answer is to change how the rear end is “laterally located”, or what transfers lateral force from the axle to the chassis. In stock form, this is done by the triangulated upper trailing arms. To a much lesser extent, the lower arms are triangulated as well. However, the rear roll center is right around the upper links, varying slightly depending on ride height.
The Cure to What Ailes You!
My solution is to remove the triangulated upper links and replace them with a single top link (i.e. a 3-link). The single top link runs parallel to the longitudinal-vertical plane of the vehicle (parallel with the longitudinal axis in a 2-D top-down view). This now frees up the axle to move laterally. You then laterally locate it with a Panhard bar or a Watt’s link. The roll center height then becomes the height of the Panhard bar at the centerline of the chassis, or the center pivot of the Watt’s link.
There is healthy debate about which lateral locating device is better, but I am going with a Panhard bar for my setup. The reason for this is two-fold. Number one is you simply can’t practically get the center pivot of a Watt’s link as low as you can get a Panhard bar. Number two is that you can actually do some tricky chassis tuning with a Panhard bar by tuning the inclination at ride height. Say a particular track/course has mostly high-speed lefts and low-speed rights. You can make the car looser around the right turns and tighter (more stable) around the lefts by changing the angle of the Panhard bar.
Yes the Panhard bar operates in an arc compared to the Watt’s link being (mostly) straight up and down as the suspension moves up/down. However, at the extremes of suspension jounce/rebound, the Panhard bar will only move the axle about 1/8″ laterally when used with coilovers.
Another very cool benefit of having a 3-link is you can offset the top link away from the centerline of the car. This has the effect of offsetting the anti-squat forces that plant the rear tires on acceleration. By offsetting the top link the correct amount, you can completely eliminate the effect of driveshaft torque on the rear axle which tends to lift the passenger tire and plant the driver tire (this is why G-body “one-wheel peels” are always on the passenger side, given equal static tire loading).
My prototype 3-Link setup has actually already been designed, simulated, and fabricated at the time of this writing. It is partially installed and I have a few more pieces to order to complete the setup. I will test it this race season, inspect after each event, make any necessary design changes, and I plan to offer this setup for sale in the Fall/Winter 2021 timeframe!