The 2G Suspension in both roll and steer
Body Roll 101
OK, first let's differentiate between static camber, camber change in bump, camber change due to castor and camber change in roll, and hopefully try to make this clear to everyone!
The 2G Suspension in both roll and steer
Now, what happens to our hypothetical T/E/L as it enters the corner? We'll assume the car is a stock 2G with about -.5deg static negative camber in front. As we enter the turn we turn the wheel. Because the top pivot of the steering knuckle is farther back than the front (this is the castor angle), the angle of the tire changes and the outside tire gets more negative camber, while the inside tire gets postitive camber, so both tires are leaning more into the turn. You can see this if you park with the wheels turned. Lets suppose in this case that the castor and the amount of steering input is enough to change the camber by .5deg. So, with our -.5deg static negative camber, our outside tire now has -1deg, and the inside tire is straight up and down.
Now the forces on the front tires change and the car deflects from it's previous course. As the tire's grip on the road tries to force the car to turn, the inertia of the car, acting thru the center of gravity (CG), will try to keep it on a straight course (Newton's law!) Because the CG is located above ground and the force on the tire is acting at ground level, the car will begin to transfer weight from the inside tires and onto the outside tires. The final weight transfer in a long corner will be proportional to the height of the CG above ground, the track (or distance between the left and right tire centers), and the weight of the vehicle. Note that even a car with NO ROLL (ie, no suspension travel) will still transfer weight to the outside tires, since the CG will always be above ground level!
Now, the outside suspension is getting more weight, and the suspension compresses. The rate at which compresses is determined by the shock's stiffenss in compression (or bump), and the spring rate, which also controls how much compression we get in the end. As the outside suspension compresses, the upper control arm, being shorter than the lower one, will bring the top of the tire in, making the camber more negative. At the same time, the suspension to the inside of the turn is uncompressed (rebounds), and the top of the tire is pushed out, increasing positive camber. So both tires again lean more into the turn. For simplicity's sake, lets assume that the suspension changes the angle of the tire 1.5deg, so the outside tire ends up 2.5deg of total camber, and the inside tire is leaning into the turn with 1.5deg of camber. This angle, however is relative to the body of the car itself, and not the the ground!
Ok, so we've got that outside tire with lots of weight on it, and leaning into the turn, right? Well, here's the bad part. As the outside suspension compresses, and the inside suspension rebounds, the car body itself rolls relative to the suspension mounting points. The car will roll about the "roll center", which is much like the center of rotation of a top. The roll center for the front suspension of our cars should be somewhere below the bottom of the engine and is probably a few inches below the ground level. Unlike a top, a car's roll center does not coincide with the center of gravity.
Remember the inertia of the car is trying to pull it in a straight line, or towards the outside of the turn? Well, since the roll center of the suspension is below the center of gravity, the top of the car is pulled towards the outside of the turn. This is "body roll" and can be pretty bad in stock DSM's. As the body rolls, the top suspension arms move towards the outside of the turn, while the lower arms remain fairly static, and this moves the top of the tire out. Now, although the outer tire is leaning into the turn relative to the body, it is leaning out relative to the car, and we're riding on the shoulder of the tire instead of keeping the contact patch flat!
So, although the geometry of the suspension gives us negative camber relative to the body of the car, the body roll of the car itself produces positive camber relative to the road on the outside tire. If you look at pictures of stock DSM's in hard turns you can see that the roll can easily use up that 2.5deg of camber on the outside tire that the suspension generated. This is particularly true of cars with R-compound tires who are generating more cornering force and hence more body roll. To make matters worse, the outside front suspension will run out of travel in hard turns, and while the body will continue to roll somewhat, the suspension can't help offset the roll anymore, and the camber will just get more and more positive. At some point, the decreasing traction of the outside front wheel will drop below the force needed to maintain the turn, and we get understeer. So the suspension gains negative camber under cornering, but still ends up with positive camber relative to the road.
Ok, now for some design considerations. I talked a lot about the outside front tire for good reason. This tire is doing the most work by far. Already, we have about 30% of the weight of the car on it just sitting still. In a hard corner, even on street tires, we can get almost 50% of the car's weight on it! Now, most tires produce the most cornering force with a slight amount of negative camber (less than .5deg), because of the flex of the sidewall (sorry, Dennis, straight up isn't always best!), but most importantly, we don't want it to lean to the outside of the turn. To keep the tire upright, we must either reduce body roll or increase the camber change of the suspension or start out with more negative camber.
Starting out with more negative camber is the easiest method, because, with a camber adjustment kit, it's simple to dial in more static negative camber. However, this increases tire wear on the street. Also, with the tire sitting without the tread fully on the road, braking and acceleration will suffer. So we can't get too carried away with the static negative camber.
We can also increase the camber change as the suspension compresses, usually by changing the relative sizes of the control arms. This gets to be expensive. Also, since the nose dives under braking, we get more negative camber and LESS contact patch, and we can't stop as quickly. So we don't want to go too far here, either. We can, however, increase the castor angle, which will result in more negative camber only when we turn the wheel. Most of the 1G camber plates also allow you to dial in more castor as well as camber, and this is a good idea. It's difficult to get too much castor in one of our cars, but too much will increase steering effort and change the angle of forces on the suspension parts.
Mostly, however, we want to reduce body roll. As we have discovered, body roll is a function of the compression of the outside suspension and the expansion of the inside suspension, as well as the relative positioning of the center of gravity and center of roll. We could change the suspension so that the center of roll is closer to the center of gravity. This would reduce body roll, but we have one problem...the center of roll doesn't behave nicely and stay in one place! As the suspension moves, the roll center moves around some. The closer to the CG it is, the more the movement of the roll center affects the car. For example, if the roll center starts 4" from the CG, and moves 2" towards it, the affect on the car is halved! This generally causes the car to behave badly and unpredictably. Note, if the CG is BELOW the roll center, the car will roll INTO the turn, instead of to the outside...you can see this with some older racing Porsches. This is hard to do, and usually results into a roll center near the CG and a skiddish car.
We can also lower the car, and therefore lower the CG and reduce the weight transfer. With less weight transfer, the suspension will move less, and the body will roll less. This is why lowering springs improve handling. We must be careful lowering the car, however, so we don't bottom out and unsettle it or run out of suspension travel and hit the bumpstops hard.
Finally, to reduce body roll, we can increase spring rate, either with springs or swaybars. As the weight transfers, stiffer springs will compress less, and so produce less body roll. Ok, so why don't we just weld the shocks or put massive springs on and get no body roll at all? Well, if you hit a bump with such a suspension, the force of the impact is fully transmitted to the body of the car. So, if the bump is 1" high, the body of the car will be deflected at least 1", and usually more because the motion of the car will continue to move it upward after it hits the bump, much like a stunt car hitting a ramp. Now, as the tire goes over the bump, it suddenly finds itself 1" above the surface of the road! Not good. We need suspension movement to keep the tire on the ground. On a smoother surface, we can go with stiffer springs and swaybars on a lower car and get much less body roll. On a rougher road, we have to soften the suspension to keep the tires in contact with the road, but we're probably not going to get as much grip anyway to generate body roll.
Now, lets consider what the Mitsu engineers were thinking. First, the car must be have enough ground clearance to go over most roads and speed bumps and not bottom out. This, of course, keeps the center of gravity fairly high. Secondly, they have to provide a fairly comfortable ride over those roads, so there is a limit on the spring rate they can use. Finally, the car must handle safely and predictably for the average driver. This means the roll center is a good distance from the CG, and the car must have final understeer in cornering, and must not have a tendancy to flip over. To acheive final understeer, the front suspension is designed to produce less negative camber in bump, so that as the car rolls more and more, the front tires lose more traction than the back. This is done by the relative lengths of the suspension arms.
Now, Dennis says that passenger cars are designed to produce positive camber in roll. This is more or less true. Since it would be impossible to design a suspension that would keep the tire straight up (or with slight negative camber) in all cases in varying road conditions and different possible tires, the suspension engineer must either make it so that it will be always gaining positive camber, or always gaining negative camber relative to the road, so to keep the suspension more predictable. So, the suspension is designed to produce NEGATIVE camber in bump, but not so much that, at the limit, the camber relative to the road is negative...In other words, it's going to go positive in almost all cases. Why? Well, imagine average Joe Driver...He's over-cooked that offramp turn, he's got the car in full roll with most of the weight on that outside front tire, and it's at the limit of traction. Now, if the tire has negative camber relative to the road and it rolls just a bit more, the tire will GAIN a bit more traction, roll a bit more, gain a bit more traction...and it will flip over the outside front wheel! I have seen a heavily modified Mustang preform this exact maneuver. Now, if the tire has positive camber with the road, and rolls a bit more, the tire loses traction and a balance in traction and roll is reached. Joe Driver may wind up in the guard rail, but at least he's upright.
Ok, so that's Body Roll 101! Any questions?