Rear Steering, not really new - Page 3 - Rennlist - Porsche Discussion Forums

Notices
991 GT3 GT3RS and 911R
Sponsored by:

Rear Steering, not really new

 
Old 03-08-2013, 03:45 AM
  #31  
Bill_C4S
User
 
Bill_C4S's Avatar
 
Join Date: May 2002
Location: London, UK
Posts: 1,134
Default

Originally Posted by axhoaxho View Post
Link no workie.

The United States of America is a little bit bigger country than UK or Switzerland. Seriously, the PSDS is in AL and I am in CA, that is more than two thousand miles away.

The US PSDS don't 'just' have a GT3 for rent. Even for the 997 GT3 Cup Experience ($10K 2-day, $3.5K 1-day), who will 'realistically' spend days of time plus so much money, travel two thousand some miles just for a 'test drive.'
Link corrected.

Actually I would ...and did....well at least to test drive the 997.1 (Masters PDE Birmingham)...now granted the test driving was an ancillary benefit...but was a motivation.

Look at the GT3 course as an opportunity to really evaluate the GT3 versus a serious benchmark (GT3 Cup) on a track..with some pretty damn good instruction....all for either 3% or <10% of the purchase price.

[And as for appreciating the broad, swathe of the US...spent a decade in Houston....four in Boston... and grew up in OZ...also as to the size of the US...i know states (Western Australia) that make Texas look small ]
Bill_C4S is offline  
Old 03-08-2013, 03:54 AM
  #32  
Bill_C4S
User
 
Bill_C4S's Avatar
 
Join Date: May 2002
Location: London, UK
Posts: 1,134
Default

Originally Posted by Mike in CA View Post
Great post Pete. Really interesting and informative. Your autocross comments, among others, were dead on. Thanks.
+1

Great to see some engineering analysis.

Ahh yes, the Weissach rear axle...quite the innovative solution in its day.

from wiki:
"The Weissach axle (pronounced 'Vise-' as in 'advise' and '-ach' with guttural 'ch') is a special rear suspension arrangement devised for the Porsche 928. The goal of the Weissach axle was to eliminate lift-throttle oversteer by allowing the rear suspension to adjust itself during cornering maneuvers.

The Weissach axle is a variant of the semi-trailing arm suspension. The tendency of a vehicle to oversteer when decelerating is compounded by the compliant bushings found in most trailing arm suspensions. When the vehicle is decelerating, the trailing arm pivots towards the rear as the wheel is "pulled" backwards relative to the chassis. This results in toe out, which makes the vehicle unstable.

For the Weissach axle, the front pivot bushing of the trailing arm is replaced by a short link. In this arrangement, when the vehicle decelerates and the wheel is "pulled" back, the result is toe in. This adds to stability and thus, reduces oversteer.

Suspensions with this or a similar function are sometimes called "passive rear steer".
Bill_C4S is offline  
Old 03-08-2013, 04:05 AM
  #33  
axhoaxho
User
 
axhoaxho's Avatar
 
Join Date: Dec 2002
Location: San Francisco Bay Area, CA
Posts: 1,485
Default

Originally Posted by Bill_C4S View Post
Link corrected.

Actually I would ...and did....well at least to test drive the 997.1 (Masters PDE Birmingham)...now granted the test driving was an ancillary benefit...but was a motivation.
Ancillary benefit...
axhoaxho is offline  
Old 03-09-2013, 12:26 PM
  #34  
SPD YLO
User
 
SPD YLO's Avatar
 
Join Date: Dec 2009
Location: SoCal
Posts: 226
Default

Originally Posted by axhoaxho View Post
who will 'realistically' spend days of time plus so much money, travel two thousand some miles just for a 'test drive.'
MJones
SPD YLO is offline  
Old 03-09-2013, 01:00 PM
  #35  
Bill_C4S
User
 
Bill_C4S's Avatar
 
Join Date: May 2002
Location: London, UK
Posts: 1,134
Default

Originally Posted by SPD YLO View Post
MJones
and not just once.....!
Bill_C4S is offline  
Old 03-09-2013, 03:42 PM
  #36  
axhoaxho
User
 
axhoaxho's Avatar
 
Join Date: Dec 2002
Location: San Francisco Bay Area, CA
Posts: 1,485
Default

Originally Posted by SPD YLO View Post
MJones
Funny.
axhoaxho is offline  
Old 03-09-2013, 07:05 PM
  #37  
Carrera GT
Wordsmith
Rennlist Member
 
Carrera GT's Avatar
 
Join Date: Feb 2003
Posts: 8,612
Default

Originally Posted by Petevb View Post
I don't know how to change lanes on a straight line, but take a GT3 through a 50+ mph box or offset slalom as fast as a Cayman can do it and you'll realize the rear of the GT3 is seriously trying to step out on the second move if you've got it set up neutrally. It's still one of the main drawbacks of the rear-engine layout.


Never. And despite that, the rear steer will still help.

As the O.O.P I'm going to try and clarify why you should consider driving the rear-steer car before making a judgement you many not yet be qualified to make.

First, I'll tell you to drive a 70s or 80s 911 at the limit and then drive a GT3. Notice that feeling when you lift off and turn in with the older 911 and the rear tries to come around? The feeling that's almost gone from the GT3? That's rear steering helping the GT3, plain and simple. It's not subtle, it's a huge effect, and the only reason the rear-engine platform is still viable today.

To get more technical, I'm going to pull some excerpts from an SAE paper, 2011-01-0987 "Investigating and Improving Vehicle Transient Handling Performance"- please look it up for a fuller explanation than I can give.

Basically rear engine cars are inherently going to try to oversteer during transitions. They have two big factors stacked against them: rear weight bias, and high polar moment.

The comparison graph below shows this, once you're able to decipher it. Below is the effect of weight distribution alone on rear tire slip angle, ie oversteer, during a "step steer" maneuver starting at one second, basically turn-in on a higher speed corner:



What you see: the larger the slip angle at the rear (6 degrees vs 3.5 peak here) the more the rear end of the car is beginning to drift. Notice how the angle overshoots as the car turns in before settling down, and also notice the size of the overshoot is greater the higher the weight on the rear axle? This is old school 911 behavior, where the rear steps out on turn-in and the driver needs to catch the rotation- any classic 911 driver who competes is familiar with it. The behavior gets magnified the higher the weight on the rear axle and the higher the polar moment, two things the 911 has in great abundance. Where you get into particular trouble is if the car turns one way, the rear adopts a slip angle, and then you need to quickly turn back the other direction- this is a recipe for a spin, and it's one of the things that defines how "neutral" one can tune the handling.

Now look at the second graph. This is what happens when you include a very rudimentary "rear steer" on the rear axle, basically bump-steer where the rear wheels turn as the body rolls:



This is turning the rear wheels slightly into the direction of the corner, and notice how much less slip angle (drift) the rear of the car has. Porsche already uses this in the GT3, but it also adds compliance steer (turns the wheels based on sideload, not position) on both the front and rear axles to neutralize the inherent oversteer of the rear engine layout much more than shown above. In a modern GT3, without you knowing it or doing anything both the front and rear are simultaneously countersteering as you enter a corner to prevent the rear end from starting to come around.

Now we know this works pretty well, so why bother adding an active system?

Two probable reasons as I see it:

First, you see significant overshoot in slip angle on the graphs, even with a little rear-steering as on the second graph. With a passive system this is probably unavoidable, as it by necessity reacts to body roll and the forces placed upon it. With an active system, however, it's possible to predict the overshoot and neutralize it entirely.

Second, the amount of this turn-in oversteer is very speed dependent. You might never notice oversteer in your car at high speed, but I guarantee you that if you've ever autocrossed you've noticed low-speed understeer in tight corners. This is easily fixable by stiffening the rear bar, but as you probably know if you do you'll take out every cone down the high speed slalom due to the oversteer effect above. Put simply you have few options in a standard car: because the transient oversteer is speed dependent you can choose which speed you want the car to be neutral at, then understeer or oversteer above and below that. Active rear steer simply corrects this issue: you just dial in as much rear steer as is needed to be neutral at any speed.

Because of the rear weight bias, rear steer is much more important in the 911 than in any Prelude or 300Z. This is not rocket science or marketing- any decent vehicle dynamics guy knows the advantages of steering the rear wheels these days (the paper noted above came out of GM), and all of the high end sports cars are using passive systems.

Just because you don't know you want active rear steer doesn't mean it's bad. My advice is to understand the system, then try it before you pass judgement. I expect it allows the car to be much sharper and faster in medium speed corners while maintaining or improving the high speed balance.
Thanks for contributing this information. I was reading it yesterday and I ended up going back to some of my notes from a former life learning about how to setup cars, specifically with track suspension. Scribbled notes from race car engineering magazine and various other clues were important steps (leaps) forward in my understanding. If only I'd studied the fundamentals more closely when I was learning to set up 993 suspension! The more I look at it, the more I like (gulp ... sorry) the more I like not only rear steer, but differential front steering. I'm reminded of many, long dinners into the wee small hours and beyond, with wine glass stains on engineering magazines and the whole dining table strewn with crude sketches of turns and radii, wheels and control arms. My sworn statement is that I recollect the consensus was to have all four wheels radially about the axis of the turn. That time will come. And soon than one might think since, not Mercedes, but Honda, has produced such a system and it's already to market in a vehicle for sale in the USA. It's called an RLX. Yes, it's an abomination of a vehicle, but the technology is elegant in its concept, simplicity and effectiveness.

https://www.youtube.com/watch?featur...DXh8zBs#t=115s

Last edited by Carrera GT; 03-12-2013 at 07:01 PM.
Carrera GT is offline  
Old 03-12-2013, 08:45 PM
  #38  
stevecolletti
User
 
stevecolletti's Avatar
 
Join Date: Apr 2009
Location: Southern California
Posts: 1,262
Default

Originally Posted by Petevb View Post
I don't know how to change lanes on a straight line, but take a GT3 through a 50+ mph box or offset slalom as fast as a Cayman can do it and you'll realize the rear of the GT3 is seriously trying to step out on the second move if you've got it set up neutrally. It's still one of the main drawbacks of the rear-engine layout.


Never. And despite that, the rear steer will still help.

As the O.O.P I'm going to try and clarify why you should consider driving the rear-steer car before making a judgement you many not yet be qualified to make.

First, I'll tell you to drive a 70s or 80s 911 at the limit and then drive a GT3. Notice that feeling when you lift off and turn in with the older 911 and the rear tries to come around? The feeling that's almost gone from the GT3? That's rear steering helping the GT3, plain and simple. It's not subtle, it's a huge effect, and the only reason the rear-engine platform is still viable today.

To get more technical, I'm going to pull some excerpts from an SAE paper, 2011-01-0987 "Investigating and Improving Vehicle Transient Handling Performance"- please look it up for a fuller explanation than I can give.

Basically rear engine cars are inherently going to try to oversteer during transitions. They have two big factors stacked against them: rear weight bias, and high polar moment.

The comparison graph below shows this, once you're able to decipher it. Below is the effect of weight distribution alone on rear tire slip angle, ie oversteer, during a "step steer" maneuver starting at one second, basically turn-in on a higher speed corner:



What you see: the larger the slip angle at the rear (6 degrees vs 3.5 peak here) the more the rear end of the car is beginning to drift. Notice how the angle overshoots as the car turns in before settling down, and also notice the size of the overshoot is greater the higher the weight on the rear axle? This is old school 911 behavior, where the rear steps out on turn-in and the driver needs to catch the rotation- any classic 911 driver who competes is familiar with it. The behavior gets magnified the higher the weight on the rear axle and the higher the polar moment, two things the 911 has in great abundance. Where you get into particular trouble is if the car turns one way, the rear adopts a slip angle, and then you need to quickly turn back the other direction- this is a recipe for a spin, and it's one of the things that defines how "neutral" one can tune the handling.

Now look at the second graph. This is what happens when you include a very rudimentary "rear steer" on the rear axle, basically bump-steer where the rear wheels turn as the body rolls:



This is turning the rear wheels slightly into the direction of the corner, and notice how much less slip angle (drift) the rear of the car has. Porsche already uses this in the GT3, but it also adds compliance steer (turns the wheels based on sideload, not position) on both the front and rear axles to neutralize the inherent oversteer of the rear engine layout much more than shown above. In a modern GT3, without you knowing it or doing anything both the front and rear are simultaneously countersteering as you enter a corner to prevent the rear end from starting to come around.

Now we know this works pretty well, so why bother adding an active system?

Two probable reasons as I see it:

First, you see significant overshoot in slip angle on the graphs, even with a little rear-steering as on the second graph. With a passive system this is probably unavoidable, as it by necessity reacts to body roll and the forces placed upon it. With an active system, however, it's possible to predict the overshoot and neutralize it entirely.

Second, the amount of this turn-in oversteer is very speed dependent. You might never notice oversteer in your car at high speed, but I guarantee you that if you've ever autocrossed you've noticed low-speed understeer in tight corners. This is easily fixable by stiffening the rear bar, but as you probably know if you do you'll take out every cone down the high speed slalom due to the oversteer effect above. Put simply you have few options in a standard car: because the transient oversteer is speed dependent you can choose which speed you want the car to be neutral at, then understeer or oversteer above and below that. Active rear steer simply corrects this issue: you just dial in as much rear steer as is needed to be neutral at any speed.

Because of the rear weight bias, rear steer is much more important in the 911 than in any Prelude or 300Z. This is not rocket science or marketing- any decent vehicle dynamics guy knows the advantages of steering the rear wheels these days (the paper noted above came out of GM), and all of the high end sports cars are using passive systems.

Just because you don't know you want active rear steer doesn't mean it's bad. My advice is to understand the system, then try it before you pass judgement. I expect it allows the car to be much sharper and faster in medium speed corners while maintaining or improving the high speed balance.
Is this just 'managed' by those big squishy bushings in the rear and alignment settings?
And if that's the case, aren't we throwing the rear steering away when we stick the monoballs and a decent alignment on it?

Thanks!
stevecolletti is offline  
Old 03-13-2013, 12:55 PM
  #39  
DER951
Addict
Rennlist Member
 
Join Date: Oct 2004
Location: Turtle Cove, Maine
Posts: 647
Default

Originally Posted by stevecolletti View Post
Is this just 'managed' by those big squishy bushings in the rear and alignment settings?
And if that's the case, aren't we throwing the rear steering away when we stick the monoballs and a decent alignment on it?

Thanks!
I, and I think many present GT3 owners, would lke to know the answer to that.
DER951 is offline  
Old 03-13-2013, 01:50 PM
  #40  
stevecolletti
User
 
stevecolletti's Avatar
 
Join Date: Apr 2009
Location: Southern California
Posts: 1,262
Default

Originally Posted by DER951 View Post
I, and I think many present GT3 owners, would like to know the answer to that.
Exactly
I couldn't come to terms with the handling on my car until I ditched the stock bushings for the monoballs. Those stock bushings (at 4500 miles) were the only bushings I've ever been able to deflect by hand (using just the bolt), and that includes 100+K mile, 'dead' MR2 bushings.

I'd actually feel better about Porsche if I knew that those soft, mushy, easily deflected bushings were there by design - as a way of injecting some passive rear steer effect - as opposed to marketing just trying to soften the car for people.

It would also explain the 'dishonest' feeling that I could never get past when pushing the car at all.
stevecolletti is offline  
Old 03-13-2013, 03:30 PM
  #41  
Petevb
Addict
Rennlist Member
 
Petevb's Avatar
 
Join Date: Oct 2003
Posts: 2,496
Default

Originally Posted by stevecolletti View Post
Is this just 'managed' by those big squishy bushings in the rear and alignment settings?
And if that's the case, aren't we throwing the rear steering away when we stick the monoballs and a decent alignment on it?
The bushing deflection is the most obvious way they manage rear steer, but they also use kinematics and flex.

In a pre-90s 911, even if all the bushings are replaced with bearings, the rear wheel will toe-out slightly and lose camber when the car corners. Realize the loads on the contact patch is over 1000 lbs even on a lightweight older car, and a deflection of even a fraction of a degree, something you couldn't see with the eye, can be felt. When the original 911 was designed, this small amount of deflection didn't matter- the springs were soft, the tires were hard, skinny and tall, and the suspension arms deflecting were only a tiny percentage of the overall movement, and lost in the noise.

When you take that same car and turn it into a track car, however, the situation changes. With stiff springs and sticky low profile tires the deflection of the rear suspension starts to become a big factor, and it's a big problem that the rear tires try to point the car in the wrong direction.

Move to a modern car and they have addressed the issue: the loads on the suspension flex the rear wheels in under loads, not out, pointing them in the right direction. This is better, but there is certainly such a thing as too much of a good thing. The amount of toe change is proportional to cornering forces, so pulling 1.5 Gs on a track car gives you 50% more toe change than pulling 1G on the street. In reality for a track car with aggressive alignment, an alert and fast reacting driver, etc, you want less toe change, not more. In particular I think the delayed reaction of rear steer can be disconcerting- a quick driver can over-correct because the rear is also correcting.

Thus for track work you need to take some of the rear-steer out. Porsche obviously thinks this- from a 997S to a GT3 to a GT3 RS, 4.0, Cup and finally RSR, more and more rear-steer is reduced through progressively eliminating rubber bushings and having the car roll less. Realize that rear-steer is still there, even with no rubber- the kinematics and deflections still toe the rear wheels in under the load, and the loads are getting progressively higher as these cars become more race-car like.

So yes, Porsche designed the soft bushings into the GT3 on purpose. When you modify this car for track work, I believe the goal isn't to eliminate rear-steer, it's to reduce it to the right amount for what you're doing. Now the hope is that the new active rear-steer can be used to reduce the bushing flex and some of the delay and uncertainty of the current GT3 rear, while still retaining the benefits of pointing the rear wheels in the right direction.

Just for reference, the rear toe deflections of some cars compared, both measured and estimated:

Last edited by Petevb; 03-13-2013 at 05:45 PM.
Petevb is offline  
Old 03-13-2013, 04:34 PM
  #42  
Mike in CA
Addict
Rennlist Member
 
Mike in CA's Avatar
 
Join Date: Jun 2001
Location: North Bay Area, CA
Posts: 11,865
Default

^^I'm learning something from your posts, Pete. Thank you.
Mike in CA is offline  
Old 03-13-2013, 04:46 PM
  #43  
stevecolletti
User
 
stevecolletti's Avatar
 
Join Date: Apr 2009
Location: Southern California
Posts: 1,262
Default

Thanks for the explanation and chart, PetervB.

My 997.1 RS is my first 911.
I've never really cared for the dynamics of a rear weight-biased car, so I really appreciate your taking the time to explain what is happening. Your explanation gives me a greater understanding of what it is that I'm feeling, and what it is that I like and dislike in a car, and why.

What is PVX on the chart?
stevecolletti is offline  
Old 03-13-2013, 05:25 PM
  #44  
Petevb
Addict
Rennlist Member
 
Petevb's Avatar
 
Join Date: Oct 2003
Posts: 2,496
Default

Originally Posted by stevecolletti View Post
What is PVX on the chart?
PVX is my car, an early 911 based race car:
https://rennlist.com/forums/porsche-...9-gt3-cup.html

It uses a semi-trailing arm rear suspension, and while it's light, simple and adjustable, it doesn't offer the benefits of a the multi-link GT3 rear.
Petevb is offline  
Old 03-13-2013, 05:43 PM
  #45  
stevecolletti
User
 
stevecolletti's Avatar
 
Join Date: Apr 2009
Location: Southern California
Posts: 1,262
Default

Very nice.
What is the suspension travel on PVX?
stevecolletti is offline  

Thread Tools
Search this Thread
Quick Reply: Rear Steering, not really new


Contact Us About Us Advertising Cookie Policy Privacy Statement Terms of Service

© 2019 MH Sub I, LLC dba Internet Brands

We are a participant in the Amazon Services LLC Associates Program, an affiliate advertising program designed to provide a means for us to earn fees by linking to Amazon.com and affiliated sites.
 
  • Ask a Question
    Get answers from community experts
Question Title:
Description:
Your question will be posted in: