Power steering feedback

[TheJonesBoy]

Todays power steering systems seem to cater to the two finger driving style by overboosting the steering assist. I personally find this discomforting at lower speeds, and flat out dangerous at highway and track speeds. A simple unintentional twitch makes a newer PS car change lanes. It is my intention to decrease the power assist of a '04 subaru imprezza.

I can only think of a few options. One is to source a new pump with lower flow, and machine a new bracket to mate with the current mounting. The second would be to install a pressure relief valve between the pump outlet and return. The latter sounds like the easier job, but I am worried about overheating the fluid. Does anyone have any suggestions? Does anyone else out there feel that manufacturers are producing cars with uncomfortable/dangerous amounts of steering assist? (or am I the only one?!)

Thanks!
-Brian

 

[patprimmer]

Adding more castor to the front suspension setting will give more steering feel, and may be an easier and more effective action.

Fooling with pump and line pressures can impact on how the steering responds at different engine speeds, and might not give a desirable effect.

If you have the resources to make a stiffer spool valve (if that's what it's called) it will make the car non conforming to the safety laws in many countries, but I expect that route will give the performance you seek.

The valve I refer to, as I understand it, is connected to the steering input shaft. It twists according to load. If the twist is sufficient, it opens a port that allows the boost to assist. As the boost allows the output shaft to catch up, the port closes and you have more effort required again. The torsional strength and the degree of twist required to open the port, together with line pressure, controls steering feel.


Regards
pat

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.

 

[TheJonesBoy]

I am familiar with the valve, but Subaru touts it manufacturing finesse with that part, quoting tolerances into the micron level. My machining resources and vehicle downtime do not enable me to go that route.

A castor adjustment would be nice, but I would like to reduce steering assist by 50-75%. I would simply disconnect the pump, but parking lots become a problem, and that is the least elegant solution.

I also know several other people with the same problem (finding their vehicle too squirrley on the highway due to ultra-sensitive PS), and I was hoping to find something that could be modified and used on other vehicles.

I cannot believe that anyone finds modern car steering systems enjoyable, espically on sporty(er) cars. I have driven older Porsches, VW's, and BMW's, and the steering response is the heart of the driving experience. Are customers today actually requesting less steering feedback, or complaining about steering effort? Being able to put a car into a high speed skid with 1 finger on the wheel seems dangerous, irregardless of what customers demand.

 

[patprimmer]

I agree, and the japanese seem to lead the field in ultra light steering. The real problem occurs if it throws a belt half way through the corner and the ultra light gets ultra heavy in a hurry.

You should still be able to get good feel with relatively light weight with extra castor. Heavy is not the real objective, feel is.

It does not cost much to try it.

Is it possible to change the length of the arm of the upright. Maybe the factory race model is different with regards to a relativly easy to replace part like an arm.

Regards
pat

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.

 

[TheJonesBoy]

I will give it a shot. Camber plates are easy to come by, and they usually have front/back adjustments for kingpin angle.

Heavy is an issue too, though. Consider a situation where you are driving on a highway, and something suprises you. Your first instinct will be to flinch. If a steering system is too light, the car will swerve violently, and one may loose control or hit a car in a neighboring lane. A system with less assist (or damping) would absorb that low energy jerk. Right now I even have trouble staying in my lane when I hit a bump, shuffle in my seat, or change a radio station. Steering effort, self centering force, and pulling all tell you important things about the road surface and your tires avaliable grip that just get lost in these systems. I guess I have too many miles and laps with manual racks.

 

[patprimmer]

Yes, but the same characteristic lets you snap on a lot of opposite lock in a real hurry if you need to. You can't have it both ways. The difference between twitchy and responsive is in the level of control the driver has of the characteristic.

Also heavy but very quick steering can be as unstable in such a situation as light but slow.

Regards
pat

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.

 

[swall]

Another potential problem with tinkering with the p/s pressure and flow is that some systems have a pressure sensor that feeds a signal to the vehicle's ECU.This is for parking/manuevering in tight situations.The signal from the pressure sensor tells the ECU to boost the engine RPM to build more flow in the p/s pump.Your scheme could totally mess that up.

 

[GregLocock]

Bear in mind that the power steering pump can be putting 5 hp or so into the fluid. If you were to add a pressure relief valve then that energy will go into heating the p/s fluid - which will not do it any good.

I don't think you can afford to drop the volumetric flow rate - otherwise you'll run into problems with catch-up. I don't think it is safe to mess too much with the torsion bar - it is critical to the straight line performance of the steering system. Having said that we do have tuning sets of them, and can't really see why one with a higher rate wouldn't do what you want.

So I guess I'm saying that there doesn't seem to be a simple method of getting what you want - and having seen many of the variable assistance schemes on the market I'd say nobody has really cracked this one, even at the OEM level, although BMW's system looks like a good stab (if rather complex), and the EPASs are good for small cars.


Cheers

Greg Locock

 

[TheJonesBoy]

I think I am going to have to fold on the idea. I am looking for a manual rack, which will solve all my problems (more or less). It may take a bit to go lock-lock, but even in tough competition rallycrossing, that doesn't happen too often. If I can live with the current 3 turn L-L rack, I may just purge the system, grease it up a bit, loop the inlet and outlet, and run it in manual mode. Anything would be better than how fidgety this thing gets above 80mph.

 

[Swinny]

There is a way to fix it at OEM level, but not many do it elegantly. It's called Variable Ratio steering. Most of the big german guys use it to some degree with their Hydraulic systems, and of course it's quite in vogue with the EPS guys to reduce motor loads at lock.

However, it's not an economical solution to a one off. It involves quite a trick process to make the steering racks, which have the variable ratio teeth on them which mesh with a normal involute pinion.

There are other ways also to achieve some on centre stiffness, and that's with various detent mechanism available. Some of the steering gear manufacturers offer solutions like this, some good, some not so. Sadly, good steering feel is hard to come by in modern cars full stop.

 

[patprimmer]

I think good feel is easier to achieve than ever before, largely thanks to the work of Arthur Bishop, but many customers see feel as a disadvantage (I think they are wrong, but their the ones with the money), so designers are forced to make cars passive to satisy marketing.

Also, back in the late 70's I saw proof that a large touring car could lap faster with correctly set up power steering than with manual steering. This was with an LT1 Camero at Mount Panorama near Sydney, Australia.

Regards
pat

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.

 

[coughdrop]

I think that you may actually have a Tire / Geometry set-up problem here.

I run a 1984 Audi Ur quattro (turbo) which was the first of the road going 4 wheel drive "supercars" and is very similar to a modern Scooby.

The car should have a real bias to driving in a straight line even with your hands off the steering, it is 4 wheel drive and if the geometry is set up correctly it should take noticeable effort to turn. Even new cars from the dealers are sometimes not set up correctly.

Firstly I would advise visiting a specialist (4 wheel drive) wheel alignment depot to get it set up, then see how it feels. (You must tell them if you want factory spec'n or otherwise.)

Secondly I would look to fit "Asimmetric tread" pattern tires. Many friends of mine and I have found these increase the high speed stability on a highway.

Thirdly, running 1 or 2 p.s.i. below spec'n in the tires helps "feel" and lateral grip in the wet particularly. This is assuming you are not too worried about a 20% decrease in tire life !

Keep it sideways !

John.

 

[GregLocock]

But the original Quattro was an understeering pig (or polite words to that effect). Understeering pigs are easier to drive safely and quite fast, but at the limit you are wasting a lot of potential grip.

I can't really see that an understeering pig is much help in defining a good steering feel.

The mods you suggest may help, but modding castor is a better approach.

If I were to characterise JonesBoy's issue it is subjectively: "On-centre feel at high speed is too light" and objectively the likely cause is "At high speed steering wheel torque vs yaw rate is too low"

So we need to increase the feedback between the lateral forces in the tyre and the steering wheel, so increasing the trail sounds like a good move, hence the castor suggestion.

Milliken's tuning charts aren't much help for this as it is really a subjective steering issue.



Cheers

Greg Locock

 

[patprimmer]

I agree with Greg.

Also, lower tyre pressure will make it heavy but less accurate. I think this guy is chasing accuracy.

To set tyre pressure for maximum grip, you need to know a lot about the compound, track, weather, wheel loads and suspension geometry. On a road car, with road tyres, I would be setting the tyre pressure at manufacturers spec or higher, on a race car with race tyres, I would be adjusting pressure to optimise tyre temps, and steering settings to optimise feel.

Regards
pat

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.

 

[coughdrop]

I agree with the engineering mod's/comments which Greg & Pat are suggesting but, I can't help feeling (excuse the pun!) that there is something currently wrong with the vehicle. Scoobies and the like should drive rigidly in a straight line, the self centering action of the PAS is also quite aggressive. I think there's something wrong with this vehicle, unless the PAS is a very different spec'n to here in the UK ?

Greg, "understeering pig" is a bit strong me thinks;-) the very early (1980) model with skinny tyres did have understeer as a characteristic but, this was soon made more neutral with set-up and wider lower tyres. Mine is set-up for the "Nordshlieffe" and has tremendously sharp turn-in with a tendency to oversteer. This is purely through set-up but using all original suspension and steering components.

The modern (road going) Scoobies & Evo's have understeer as a characteristic to allow "front wheel drive drivers" to easily get used to the car.

Cheers,
Keep on keeping it sideways !
John.

 

[TheJonesBoy]

My two issues were this.
1) Steering effort is too low; small, unintended muscle movements get transformed into unintentional steering inputs. This is more noticeable at high speeds (80-140 mph). Torque vs yaw is too low is spot on.
2) Steering does not provide enough feedback; traction limits are hard to judge.

For instance, I was traveling down a snow covered road with a hardpack center crown and soft, unpacked sides. On my manual rack car (86 VW Jetta, understeering pig king, prior to major suspension mods), when I got close to the edge of the right hard pack, I could feel the self-centering disappear. As the right side tires entered the softpack, the car would pull with increasing torque proportional to the depth of the snow. This is from wheel slip (hardpack side) and the difficulty of the tire plowing the snow aside (soft side). This is your sign to move left. If you do not do not countersteer and gas it, the right side will be pulled hard enough to put you in an uncontrollable, violent skid. With the power steering, the only indication of danger is the power loss, which the viscous diffs mask a bit, making it feel like deep snow all around. Other situations, such as hydroplaning and ice, (espically on single side cases)can be felt as steering pulls, increased/decreased steering effort, and loss of self-centering on manual racks. Even on a dry pave sweeper, when steering effort begins to drop, you know you are approaching your traction limit and should lift/countersteer.

Adding castor would definately help problem #1 with minimal side affects (Thanks for all the input, Pat!). Rallye tires with harder sidewalls will help too, but the car needs a little prep work before I begin the dirt road drivers-ed beatings.

On a side note, I think the understeer to oversteer transition is smooth, predictable, and easy to control through a turn. Heck, I made three confident and smooth (and low speed) drift turns on my snowy test drive before the dealer freaked out and made me turn around.

John: Sideways, airborne... as long as the shiny side is up!

-Brian

 

[patprimmer]

Steering pull rather than weight is dependant on scrub radii and can be altered with change in wheel offset, or with spacers. The use of thick spacers has an obvious detrimental effect wheel studs and on front suspension durability

Regards
pat pprimmer@acay.com.au
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.

 

[JayMaechtlen]

On my last car (don't laugh, it was a '91 Buick Century)
I noticed some of the effects that Brian is complaining about. They were due to a pair of tires I installed, which had very different behaviour than the stock type tires.
When the lower wider tires were on the front, it was twitchy in a straight line. Strong turn-in, nice feel in a corner. On the rear (skinnier tires in front) it was back to docile straight-line machine.
Yes, it is possiblle there were other problems. It was just a car, and I put a bunch of miles on it!
regards
Jay

Jay Maechtlen