Steering Clear Of Torque
I have this friend; we'll call him Chad, since that's his name. I get e-mails from him on a weekly basis, sometimes daily, asking about some concept he just read about or idea he's come up with. He works in the financial industry, but has a mechanical mind. I look forward to his mails because when he asks how something works, he's looking for something like: "Well, a limited-slip differential has worm gears, so they'll move freely with light loads applied. But apply too much load, the gears bind and they no longer spin so easily, kinda like an adjustable pipe wrench..." He always gets it after analyzing a real example he has dealt with.
During a recent discussion on car dynamics, the topic of torque steer came up. Everyone knows that torque steer is what happens when there's a lot of torque in a frontwheel- drive car. But that's like saying a hamburger is what happens when a cow gets too near to a bag of charcoal. There are a few important steps in between that have been woefully ignored.
First, torque steer is caused by powering the steered wheels on a vehicle. All steered wheels have what's called a scrub radius. The kingpin axis, the line about which the wheel is pivoted when steering, is near vertical and can be visualized by drawing a line between the upper strut mount and the outer ball joint on a MacPherson-equipped car. The scrub radius is the distance from where that line intersects the ground to the center of pressure of the associated tire's contact patch. A center of pressure located inboard of the steering axis is considered negative; outboard is positive. It should be noted that some cars have zero static scrub radius.
When torque is applied to the road, the tire wants to pivot around the kingpin axis. If both sides are putting the same amount of torque to the ground, they balance each other out. If one slips and one grips, then the imbalance causes torque steer. But that isn't the whole story, or any car equipped with a limited-slip differential wouldn't have torque steer.
Torque steer can also be caused by the drivetrain itself. It's a misconception that unequal-length axles are the major cause of torque steer-that's only partially true. Unequallength axles cause the CV joints to operate at different angles from one side of the car to the other. Whenever torque is applied to an object that isn't completely in line with the object's axis of rotation, it wants to turn the object in the direction of the torque instead. The more extreme the angle that the CV joint has to make, the more torque goes to steering the wheel as opposed to driving it.
Can torque steer be eliminated? Yes, but it's tough to accomplish without some sacrifices. A car can be built with zero scrub radius, which isn't that complicated, but it'll lose steering feel and increase steering effort at low speeds. Cars could be designed with equal-length driveshafts that level with the hubs. These have become relatively common, but it becomes next to impossible to keep them completely horizontal because manufacturers have to account for variations in ride height and load, plus enthusiasts have a habit of changing the ride height on their vehicles whether they need to or not. Limited-slip differentials are also becoming more prevalent as factory equipment, but an overly aggressive differential can compromise steering feel and, again, make low-speed cornering difficult.
One obvious solution is just to live with it. Unfortunately, I'm guessing that somewhere there's an engineer with too much free time designing a computercontrolled active anti-torque steering system. It will weigh 200 pounds, add a $1,000 to the overall cost and take away even more driver control. Given enough time and e-mails, I'm also guessing Chad and I will MacGyver a mechanical system using pie tins and bungee cords. Until that happens, just use the steering wheel.