Please explain "torque steer"

February 25, 2004 5:56 PM

Torque steer- I'm an engineer, but I've never had torque steer adequately explained to me. Most mechanics will tell you that it's due to unequal length axles, but after the initial loading, the greater twist in the longer axle shouldn't count. Anyone?
posted by notsnot to science & nature (6 comments total)

I thought torque steer was a uniquely front-wheel drive phenomenon resulting from the inherent instability that results from trying to steer with the same wheels that drive the vehicle forward.

But maybe you're talking about something else entirely.
posted by trharlan at 6:25 PM on February 25, 2004

Unique to fwd vehicles with unequal length axles. Subarus don't have torque steer: the axles are the same length.

My little NX 1600, on the other hand, torques like a bugger when you haul ass off the stop line.

The steering wheel actually twists in your hands. Naturally, you fight back for the straight line. It's a momentary thing, less than a tenth of a second, and then everything evens out.

The long axle twists, letting the short-side wheel get the jump on the long-side one. This also seems to shift the rack (of rack & pinion steering) position, twisting the steering wheel.

It all untwists, of course. Because you maintain the straight line, the long-side wheel needs to catch-up to the short-side one. There's only so much torque available to twist the thing, and you aren't in full-power output for long, so the twist can ground itself out easily enough in the end.

I suppose if you had unlimited torque power, you'd eventually snap the axle; or the twist would empty itself out through the transmission linkage and into the short-side wheel...
posted by five fresh fish at 7:10 PM on February 25, 2004

See, that's just it - the long-short axle answer doesn't satisfy my brain--it seems like there ought to be more to it. If the short-axle side is getting the jump, maybe five degrees, that *should* be it, right? After that five-degree difference balances, there shouldn't be any more torque steer. I know that torque steer can be removed by special suspension-alignment tricks, so I guess I thought there was something to do with suspension-load force oblique to the gyroscopic motion of the motor creating the steer. (like steering a motorcycle left in the yaw direction against the wheel which is turning in the pitch direction making the whole bike roll to the right...)
posted by notsnot at 9:04 PM on February 25, 2004

Torque steer lasts only as long as the engine can outpower the axle. As soon as the axle hits maximum twist the show is over. Long-side wheel begins to rotate, soon catches up to the short-side wheel.
posted by five fresh fish at 2:15 AM on February 26, 2004

> If the short-axle side is getting the jump, maybe five
> degrees, that *should* be it, right? After that five-degree
> difference balances, there shouldn't be any more torque steer.

When the difference balances, the torque steer does go away. It only happens under initial acceleration, while engine power is sufficient to maintain the twist differential between the short and long half-shafts. It all works out just the way you think it should.
posted by jfuller at 6:34 AM on February 26, 2004

i think you're right in suspecting the axle length thing. it doesn't make any sense to me - do people really think an axle twists up like wringing a wet cloth?

anyway, i went googling, and we're not alone in doubting the explanation. there's even people researching it. and the best explanation i've found is here (which gives the basic mechanism, and is open and honest (some sites read like the authors don't understand the long words, but hop they'll scare people into believing them - this one doesn't), but still doesn't provide a clear answer about why it would be systematically in one direction).

incidentally, the first link i give, above, shows how one manufacturer corrected for this problem by making the axles equal length. only they're not equal length - the longer one is still longer, but now has an extra joint. that would make it less rigid, i think. more likely to my mind it makes the geometry at both wheels the same no matter what the suspension level - that keeps the emphasis on asymmetry but doesn't require axles that wind up like rubber.
posted by andrew cooke at 9:32 AM on February 26, 2004