Torque Steer

[dcroasmu]

I've been looking high and low and haven't been able to find a satisfactory explanation of torque steer in front wheel drive vehicles. I understand that it's caused by unequal length shafts from the differential to the drive wheels, but I don't understand how unequal torques come out of an open differential.

If anyone knows where I can find an in-depth explanation, I could stop pulling my hair out. Thanks,

Dan

 

[diskullman]

Funny, I get torque steer in my front wheel drive car, even though the shafts are equal length. Maybe taking a ride in my new Impala SS, with 303 HP driving the front wheels will give you a "hands-on" understanding of torque steer, as when you mash the pedal from a dead stop, it wants to drive off into the woods. Hold on to the wheel, TIGHT!
As far as I know (I'm not an engineer) the torque is transmitted to the wheel that grips, not both equally.
best definition I can find is at

http://www.sportcompactcarweb.com/editors/technobabble/9909scc_technobabble/

Russell Giuliano
Unique Technologies Associates - Cobra Solid Lubricant

http://www.uniquetechnologies.com

 

[Warpspeed]

O/k, I will have a go at this.

My understanding is that torque steer effect mainly has to do with the caster and scrub radius built into the front suspension rather than having different drive shaft lengths.

A certain amount of caster will cause the steering to self centre and provide some tactile steering "feel" and driver feedback. These forces increase during braking, improving straight line stability. Unfortunately during acceleration with FWD it all works backwards, the more wheel torque you have the more unstable it can become.

Scrub radius will generate a steering torque on each side of the car. Provided wheel torque and weight is the same on each front wheel the forces should balance, with no net steering effect. During braking, caster should be sufficient to ensure the whole mess does not become unstable. But during acceleration it may start to steer, and lateral weight transfer can exaggerate the effect, so it may then become very difficult to hold a straight line.

I know my turbocharged FWD would gently rock from side to side and weave with ever increasing amplitude under high acceleration. Rather disconcerting, and it had equal length front driveshafts too.

 

[mloew]

Go back to first principles. Resolve all the forces and moments acting on the system (wheel end) around the steering axis. This resultant torque can then be further resolved into the forces acting on the steering tie-rods at each side. The net force on the steering system connecting the two sides results in "torque steer". This can come from several sources: resultant torque from half shafts operating at angles, different normal or longitudinal forces, etc. Many questions like this can be answered by taking a good look at the Free Body Diagram.

Best regards,

Matthew Ian Loew


Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[IceStationZebra]

I have also heard the unequal length drive shaft explanation. I have two reasons that this might be valid, though I can't verify either.

1. The first reason could be different drive shaft angles. On early designs the drive shafts usually went directly from the transmission to the spindles and the differential was not centered. So depending on the ride height the drive shafts would operate at different angles. I know that CV joints shouldn't cause this, but maybe the different angles create different loads on the spindles? Do CV joints have significant differences in friction with a slight change in drive angle? 3 ball vs. 5 ball? I supose that this would also be valid on a car with equal length shafts if the car was not sitting level side-to-side.

2. Before I heard the different angle idea, my thought was that the unequal length drive shafts would have different amounts of wind-up on the left vs right. ie different shaft diameters, lengths, etc. And even if the drive shafts are equal, unless the differential is in the center of the car there is still a jack shaft on the right side which would cause a difference in stiffness.


I had a 92 Old Achieva with the 190hp Quad four, Getrag torque biased differential, and equal length drive shafts. It could spin the tires in 2nd gear and it would yank the wheel for a split second until the diff kicked in. It always pulled to the same side, right as I recall. This would lead me to believe it has something to do with the diff, but I can't fathom a reason.

Happy hunting, ISZ

 

[GregLocock]

Even cars with equal length halfshafts suffer from torque steer. I am not sure that anybody has a really good understanding of all the reasons.

Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[mloew]

Greg,

I have a very good understanding. Statics gets you most of the way there. I really wish people would stop guessing and offering opinions and instead frame the question in terms of engineering principles. Sad to see the FBD so under-appreciated.

Best regards,

Matthew Ian Loew


Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[modesign]

Two subjects relevant to the last postings from Greg and M Loew. First in my experience as a successful designer of mechanical devices I have found educated guesses and relevant opinions very valuable. If all problems could be solved by the use of FBD why do some cars have bad torque steer and others not? Why with the use of every computer simulation and almost indefinite money supply on all parts of the top GP cars is one better? Gregs point that we do not always understand ALL the reasons for some subject is well made. In any case guesses and opinions generate replies that are educational!
Secondly whatever the reason for it,torque steer effect felt by the driver is very dependent on the road surface, camber and undulations as with torque steer the car tends to follow the down side. Opinion based on driving a number of cars on a variety of cambers.
Regards
Sandy Cormack

 

[GregLocock]

OK Mathew, you are right, I haven't drawn the FBDs, because the puzzling case is one for a symmetrical suspension.

However, thinking about it this morning I came up with a list of assymetries that reliably exist in the real world.

I'm going to be a bit hazy about this as I want to see if the numbers stack up.

Incidentally it is probably worth pointing out that classic torque steer is a FWD problem, and to some extent is related to the steering wheel torque and steering system compliance, that is, I think that if the driver held the road wheels straight then the car would not steer any more than a RWD.





Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[patprimmer]

I am sticking my neck out here (or at least more than usual) as I have no data and I am only reacting to casual observation and intuition, but I find torque steer seems very susceptible to minor changes in settings and surface, so I suspect the true centre of force for the tyre might not be the centre of the contact patch and might move sideways from the centre with camber change, or with road surface change, and back and forth with caster change. If the effective scrub radii moves even slightly at a different rate from side to side or the trail moves back and forth, it would produce the torque steer effect.

Regards

eng-tips, by professional engineers for professional engineers
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[GregLocock]

I think that and road camber are part of the answer, but I am pretty sure that some models of car always torque steered the same way, even with notionally symmetrical drivelines and suspensions.



Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[NormPeterson]

Is it possible that the engine/transaxle rotates a bit about the X-axis in response to traction forces? Compliant bushings do allow motions that simplified pin-connected models do not.

Norm

 

[GregLocock]

They do, because engine mounting systems are rarely optimised to prevent it completely. With three mounts it is possible to change the ratio of the stiffnesses so that the engine only pitches, but in practice that can compromise your isolation. The maths is straightforward, the practical details are not.

Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[NormPeterson]

So it's possible for equal length halfshafts to lose the static symmetry of joint angles via the combination of powertrain rotation about the X axis and the nose rising due to longitudinal load transfer.

Is this phenomenon perhaps more prevalent in models for which NVH isolation and ride comfort carry relatively higher priority?

FWIW, and not being a drag racer, I've noticed this only a couple of times in my '01 Nissan Maxima and never in the '95 Mazda 626 (which does experience quite enthusiastic starts at autocross). Both have manual shift transmissions, though the Maxima also has a viscous limited slip diff.

Norm

 

[GregLocock]

"So it's possible for equal length halfshafts to lose the static symmetry of joint angles via the combination of powertrain rotation about the X axis and the nose rising due to longitudinal load transfer."

Yup. Well, that's another one to put on the list.

"Is this phenomenon perhaps more prevalent in models for which NVH isolation and ride comfort carry relatively higher priority?"

And/Or maybe in those manufacturers who don't understand how to mount FWD engines properly?

Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[evelrod]

Well, y'all know my little racer has phenominal torque steer, especially at very low speed in the paddock...on track, at speed, it's not noticable. That's just 'FYI'...the thing I don't see mentioned here is the balance of front corner weights...in my little Mini, that's one of THE most important settings!

Rod

 

[Warpspeed]

I have found the same thing. Corner weights, and having the same caster settings on both sides go a long way to reducing torque steer.

 

[eddyesi]

couple of other reasons missed

1) outer CV joints do create a 2nd turning moment due to angle. This was indicated with the FBD discussion but not expanded on. When at different angles (unequal length) this creates a difference in turning moment accross the axle which is proportional to torque, so you get torque steer

you still get this on equal length shafts due to differences in angle from roll, bump and steering (akerman etc), so there is no such thing as a FWD with no torque steer

2) there is vary rarely such a thing as 'equal length' driveshafts in FWD cars, as even though the barshafts between CV joints can be equal, and is aimed for for the reasons above, the total length isnt, as the differential is rarely central, so a linkshaft is used on one side. This creates a difference in torsional stifness betwwen L and R shafts, so during torque dynamics this results in differences in wheel end torques due to the shaft windup, and therefore torque steer. This can be reduced by trying to equalise the torsional stifness also, but is often difficult to achieve to <10% difference

3) You also have to look at suspension geometry and things like variation in bush rates and deflection, which also have an influence

the aim is to achieve a suspension and steering system that is largeley insensitive to torque steer, as well as minimise the driveline effect on it, as there is no real way to ensure equal wheel torque across the axle

If you want a good example of a FWD suspension system that is insensitive to torque steer, look at Hondas 5th gen Prelude fitted with ATTS (active torque transfer) diff. This had a unique unequal length independant suspension at the front end with double control arm links.

If you want some bad examples look ad the Ford Focus RS and Rover 220 turbo (Euro models). Both of these had McPhereson front suspension (wich is far from ideal) with Torsen differentials and high torque FWD powertrains, and both had major issues with torque steer sensitivity!