Page 27 UNDER/OVERSTEER, ROLL UNDER/OVERSTEER, CAR STABILITY, AND STABILITY OF THE CAR/DRIVER SYSTEM
The earliest definition of understeer and oversteer is widely attributed to Maurice Olley. It can be demonstrated in the characteristics of a car as it is driven on a crowned road, such as was common in Olley's day. If, when the steering wheel is in the straight ahead position, the transverse gravitational pull tends to cause the car to run off into the ditch, it is understeering. If, instead, it tends to turn back into oncoming traffic, it is oversteering.
The more modern SAE definition includes the steering wheel angle as a parameter, but both definitions will identify the applicable characteristic in a given car.
The "critical speed" is a significant measure of a car's stability. If operation is attempted above this speed, the slightest steering wheel input will cause loss of control. Fortunately, the critical speed is infinite for an understeering car. For an oversteering car, however, the critical speed is finite and has a lower value as the amount of oversteer is increased.
As the discussion of tire loadings...on another page at this site...would indicate, the existence of under or oversteer (excluding, as we shall see, that which is called "roll over/understeer") ultimately is determined by the loadings at the four tire patches.
The driver recognizes the presence of oversteer by the "extra" steering wheel inputs required. Even in the simplest of maneuvers, he finds himself "sawing" the steering wheel back and forth.
Roll steer effects can occur with an IRS or with a beam rear axle, but I'm going to confine my comments here to the beam axle. When attempting to control rear wheel dynamic loads during the lauch of a dragracing car, suspension geometries often result in a significant amount of roll oversteer. When analyzing the behavior of a car with roll oversteer, it will be found that both Olley's test and the SAE definition would identify the car as "oversteering." It does not follow, however, that a car with significant roll oversteer has a critical speed. This might seem a contradiction, but, when we consider the tire loadings, the contradiction disappears.
Both the Olley test and the SAE definition involve steering wheel angle. In the case of the Olley test, the steering angle is to remain zero. This angle is measured relative to the car's centerline (SAE X axis). But, this is arbitrary and assumes that the axle centerline remains perpendicular to the car's centerline. If, instead, we measured the steering angle relative to a line perpendicular to the axle centerline (and kept this value equal to zero), the car would...if it had been understeering before the addition of the roll oversteer...head into the ditch. This is because the tire patch loadings have not been affected by the roll oversteer.
So, as far as the car's stability is concerned, we can add roll oversteer without effect. It is only necessary that a change in steering input be made to compensate for the rotation of the axle assembly. (Actually, this isn't completely true, for there are transient effects as the car enters and exits the corner. The car must be rotated about the vertical (Z) axis and roll steer, whether under or over, changes the total amount of initial rotation. The inertial effect must be taken by the tires. The result, with a roll oversteer car turning left, would be a clockwise torque, on the four tire patches, as the car enters the corner and a counter-clockwise torque as it exits. Again, this is a transient effect, occuring within a very small time frame.)
But, before we get carried away, we must realize that any practical consideration of stability must encompass the driver. We must consider the car/driver system. All of those "extra" steering inputs must be made with a roll oversteer car, just as with a car that oversteers due to the nature of the tire patch loads. This is certainly not a "driver friendly" environment, to put it mildly. The results, indeed, can be accurately identified as an instability.
The preceding should not deter one from attempting cancellation of the driveshaft torque in a dragrace car. Again, as far as the car is concerned, there is no basic stability problem. The driver must, however, recognize that a bit more effort is involved as far as steering input is concerned.