Objectively measuring and defining subjective steering "feel"

[vcebrad]

I recently read an article on the new C6 Corvette and noticed GM got high marks for many things, but steering "feel" was not one of them. Many of the BMW's, especially the previous (mid 90's?) model M3, are widely touted as having nearly perfect steering.

What is it about steering system design parameters that make that tiny difference and push a great car into the realm of the nearly perfect? Are there books out there on automotive steering system design? Are there experts who make fortunes at BMW, know enough to write the books, and are impossible for GM to lure away? How does one objectively measure and design in a certain steering feel, when it seems to be a subjective criteria?

 

[evelrod]

Feel? There is a fairly wide range I would guess. Somewhere between my "feel" as an over the hill race car driver ( I found the BMW's, ALL of them, to be seriosly overboosted) and my wife's opinion that, based on her 'feel', they were just right. I have found, with several cars and trucks I've driven lately (including some pretty high dollar offerings) to suffer from an overboosted power steering. Maybe that's my opinion but, I hear the same complaint from some of my racer buds, too. Keep in mind that I drive a 1930 Model A Ford as a 'daily driver' so I am probably not in the OEM demographic on this issue.

An aside...the seriously overboosted brake servos of yesteryear seem to be dramatically improved! ??? Maybe I'm just getting old and lazy? Even my race car now has a servo!

Rod

 

[vcebrad]

Along with the steering "heft", I think there are some other characteristics that are sought after like
- feedback
- discernible play
- accuracy
- path accuracy

Others?

 

[Joest]

I understand that Ford studied the BMW design extensively when they designed the front suspension of the new Mustang. Having recently looked at the new mustang front I would note that the spindle centerline is well ahead of the steering axis (toward the front of the car). I measured it to be about 20-25mm forward. This is interesting because doing this reduces the amount of mechanical trail. This, at first, seems to be headed in the wrong direction because modern low-profile performance tire need more mechanical trail. It could be that the design uses a very high positive caster angle that compensates for spindle's forward arrangement resulting in a net amount of mechanical trail. The forward-shifted spindle axis is probably a way to "tune" the steering feel so that it doesn't become too heavy. From my experience with racecars, the amount of (+) caster typically desired is about numerically equal to the steering axis inclination, which can easily be 15-20 degrees on a strut front suspension as the BMW.

Anyone else have additional specifics or analysis they care to contribute?

 

[GregLocock]

We have a two week practical course in steering evaluation, if that helps. I haven't been on it. It's an interesting area to study, we can measure the steering wheel torque and angle (SWT and SWA) during a manouevre, but once you've got the data traces there doesn't seem to be much rhyme or reason to any relationship and the subjective feel of the car. So in practice we measure everything possible on cars that we like, and then try and make ours like that. I'd say we have a 50% hit rate.

BMW's new EPAS is getting bagged for poor steering feel, so I guess even BMW don't always get it right. They also made some slightly odd trade-offs on the X5 (apply full lock when parking. release the wheel, for example).

There are two sides to it, there's eliminating all the wrong things (friction and lash and other non linearities for example), and then there's doing the right things, which can be split into things we know about, and all the rest. Rumsford's philosophy applies, and there's an awful lot of unknowable unknowns around.

Important objective measures include frequency response of the yaw velocity/ SWA curve, and linearity of that curve against amplitude and speed, and the shape of the loop you get when you when you plot SWT vs SWA for a gentle weave.

You should also be aware that different parts of the handling spectrum have different requirements. On centre feel needs completely different evaluation techniques, and hardware mods, to limit handling.

And as one seasoned campaigner pointed out to me, 70% of the model on model improvement is down to the tires.

For what it is worth, in my whole car dynamics model, 50% of the code is tires, and 25% is the model of the steering system. That leaves just 25% for everything else. If you ask me, the worst (least representative) part of the model is the tires, and the next worst is a tie between the shock absorbers and the rack/PAS system.



Cheers

Greg Locock

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

 

[Joest]

When the average driver report that a car has "good steering feel" I don't think they are only talking about steering feedback and precision, but also how well a car turns in (goes where you point it). Several factor are responsible for this, namely stiff springs, but a high roll center can also give this "feel". Therefore, I could conclude that "good steering feel" is a loaded work that also has implication response and ultimately handling. Most consumers find it difficult to separate these because they are so closely coupled.

-Joest

 

[GregLocock]

I agree, high RCH is a nice way of getting force into the lateral links, so it is likely to improve feel. Slightly more generally, initial roll control also has a big influence.





Cheers

Greg Locock

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

 

[hpfreak]

IMO much of the steering feel is system compliance. Focusing on rack bushings, tire sidewalls and response, control arm bushings, body/frame stiffness and weight distribution, not just boost and ratio will round out the package. Minimizing compliance goes a long way to helping steering, but must be balanced with linearity and high speed stability. Torsion bar stiffness and valve design are the typical focus, but ratio and the items mentioned are the foundation of building a great system.

BMW seems to be able to balance ride/handling better than other companies. Maybe because they set the priorities from the beginning to have "The ultimate driving experience" not "Beat Viper". Design of the frame and subframes, mounting attachment stiffness and polar inertia are things that are hard to tune.

As for measurement, steering feel is just that. Feel. Different demographics have different priorities. Tuning is done by the seat of the pants and maybe some instrumentation to create arguments later. Vehicle dynamics engineers have to put their mind in "customer mode" and try to figure out what a typical corvette buyer wants from the steering. As soon as a magazine writer gets in one, they want razor sharp response and "heft". They typically drive tracks or short term, which may justify racier steering. More conservative daily driver types, may want lighter effort to park without hitting a curb and scratching those nice rims. Again, it's balance which takes alot of tuning time to sort out interactions and find the right combination of stiffness and comfort. BMW seems to have the edge.

 

[modesign]

The BMW is rear drive most others front. Is it easier to make a good steering feel without front drive problems? (see torque steer).The point about the diffeence between the road tester on track wanting a pointy high feedback steering and the average pwerson who is more comfortable with a deader feel and possibly higher boost is very true.Audi steering/suspension is often criticized for lack of feel but around town and a reasonable speed is 'comfortable' to drive.

Regards

Sandy Cormack

 

[Tmoose]

The "debate" of FWD VS RWD goes on.

My feeling is that before the mid 60s the only tools for good "handling" were light weight, rearward weight bias, rack and pinion steering, reasonably balanced front and rear roll stiffness, and radial tires. There were way more clumsy, ponderous RWD cars than good "handlers."

It took the trans-am series to get Ford and Chevy to understand thoughtful camber curves, and to re-design the Mustang and Camaro suspension to incorporate them.

Nowadays every vocal owner of a RWD car is pretty snooty about how bad FWD "feels." When the Renault Fuego (FWD with the longitudinal engine hung out the front) first came out with nice feeling fast power steering, exotic Michelin tires, etc, Road and Track reported it had a real impressive slalom speed.

A Mustang with 54% of the its weight on the front end with nicely balanced roll stiffness may delight its owner/driver in dry weather. That's fine. Body Piercing, Cosmetic surgery and "reality" shows populated with young folks with snippy attitudes are pretty popular these days too.

Around 1983 an acquaintance who campaigned a Lotus Elan in SCCA for years was VERY enthusiastic about the VW GTI's street entertainment value. I asked if it was as fun as his Elan, and he said no, but.......
I bought a new 1984 GTI. Hey, with a responsive engine and steering, pretty decent tires, it was very entertaining, and very well balanced when cornering hard enough to set off the oil pressure light and buzzer. A friend described it like a frisky pony.

Empty, In the snow, my mom's Accord and similar vintage Camrys are almost precariously balanced between trailing throttle oversteer and understeer.

From time to time FWDrivers have ruled some and rallying and SCCA stock classes (Integra and Saab).

Ice racing classes often have separate FWD classes. I don't think that is to make the FWDrivers feel better by not having to compete with the superior RWD.

 

[jag27]

what is wrong with BMW EPAS

which one is the best performing EPAS currently?

 

[NormPeterson]

The BMW is rear drive most others front. Is it easier to make a good steering feel without front drive problems?

I would think that not having the moment effect of traction forces about the steering axis would at least simplify things (if not permit a wider range of compromises).

Recollection of my first experience driving a FWD car included "stiffening up" of the steering under power such that one had to physically steer it back straight. Fortunately, things have improved considerably since, though there are still noticeable differences between different FWD cars that are probably the result of philosophical differences among mfrs. For example, turn-in with my Mazda 626 is better than that of the Maxima by more than can be explained by differences in weight and tire size.

Norm

 

[GregLocock]

Also handling, if not steering feel, on the original Mini was fine.

And of course the Lotus Elan II did manage exceptional handling and reasonable steering feel.

The GTi has always had fairly bad steering feel, in my book. Quite why anyone would use it in this thread is beyond me, unless it was to define what lifeless, heavy, innaccurate feel is all about. They might as well have stuck 100% power assist on it and used a spring to tell you where the middle was.





Cheers

Greg Locock

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

 

[evelrod]

If finding 'center' is a problem, I have a cure...

Take one moderate to high hp FWD, add limited slip differential, mix in about 5 degrees of positive caster, select highest reduction gear, remove hands from steering wheel and apply power...voila!...instant center.

Rod ;-)

 

[bobqzzi04]

Greg,

Are you talking about the original 1984 VW Rabbit GTI? If so, I'm surprised to say the least- my subjecitve opinion (is there any other kind?) is that they had excellent steering feel. I wonder if they came spec'd with differnet tires where you are.

 

[GregLocock]

I am trying to be rigorous. Steering feel is the feedback you get via the steering wheel. I am not saying that the GTi was a bad /handling/ car, merely that the feedback via the wheel was not especially helpful.

To be honest, this is an area where RWD snobbery does have a place, a RWD car can have all sorts of nastiness in its front suspension, and it will have good steering feel, whereas an AWD/FWD will almsot always compromise its steering feel.

Cheers

Greg Locock

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

 

[Tmoose]

"RWD snobbery does have a place.... a RWD car can have all sorts of nastiness in its front suspension....it WILL have good steering feel."

I repectfully request opinions on the steering "feel" of these examples - any GM or FORD, or even MOPAR full sized or intermediate vehicle with power steering made between 1955 and 1978. I think there are a few RWDs in there.

 

[evelrod]

You gotta be kidding!

Rod

 

[GregLocock]

Shrugs.

Bad examples prove little. The market apparently wanted effortless steering, so that's what they got.



Cheers

Greg Locock

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

 

[Tmoose]

What I >>thought<< I heard was "RWD is good" and "FWD is not so good." But, after thinking about it a while, and waiting a few days, and thinking about it some more before responding, based on counter-replies from folks I respect I now believe that when I listen to FWD/RWD steering or handling questions I am simply not hearing it correctly.
I'm sorry.