Gen 2:
Alison's 2nd Generation Setups

In the Real World: Setting up the Cobra
Does Real World Setup Theory Translate to GPL?
What's Different About These Setups?
When Stiffer is Softer
Download the Setups
Adapting a Setup to Another Track
Update: F2 Setups for Lime Rock

Welcome to the next generation of Eagle Woman setups!

In the Real World: Setting up the Cobra

In the summer of 2000, our enthusiasm for racing fired by years of involvement with GPL and earlier racing sims, my brother and I built a Cobra replica.

While the Cobra, as built from the kit, was a fairly impressive-looking and -sounding street car, we soon found that in a competition environment (in other words, on race tracks) it was severely lacking in handling and braking.

This triggered a lengthy (and expensive) development process, which eventually resulted in our replacing almost every part of the original suspension and braking systems. As of July 2002, we have upgraded all four springs, all four shocks, all four uprights (spindles), all of the front control arms, the steering rack, the front brake calipers and rotors, the master cylinder, the brake pad material, and all flexible brake lines.

In addition, we've engineered and built adjustable front and rear anti-roll bars (the kit manufacturer doesn't supply any anti-roll bars!) and designed a modification to the rear uprights that eliminates all flexible (rubber or urethane) bushings in favor of teflon-lined spherical bearings.

All this was done under the guidance of famed GPL beta tester Doug Arnao, who also happens to build racing Porsches for a living. Drawing from his Porsche experience - and using his computer suspension development tools - Doug advised us on spring and shock rates (and supplied the necessary parts) and helped us design our anti-roll bars and rear upright spherical bearing conversion.

The Cobra suspension and brake modifications have taken a year and a half to accomplish, and a lot of hard work and quite a bit of money. But it was worth it. The car has been transformed.

The Cobra started out feeling marshmallow-soft, darty, wallowy and unpredictable, with mushy and fade-prone brakes. Every control input caused it to roll or pitch dramatically, and it was so under-damped that every control input and every bump caused a series of rebounds. It also tried to steer it self every time it encountered a bump or ripple in the pavement.

No matter what you asked it to do, the stock Cobra was always in your face, making you adapt to its little idiosyncrasies and foibles, forcing you to pay attention to staying on top of the car rather than what you were trying to make it do on the track.

At Watkins Glen in July 2002 - after two years of development - the Cobra behaved like a proper race car. It's now fast and stable, neutral and precise, with prodigious stopping power and ferocious grip. All of its vices are gone.

In a way, the car itself has receded from the experience of driving it. It no longer calls attention to itself; it just goes where you point it and does what you tell it. It's a joy to drive at speed or on the street.

Does Real World Setup Theory Translate to GPL?

Although there were a lot of things wrong with the Cobra that aren't wrong with GPL's cars, there are also many important aspects of the Cobra's setup that apply to any race car. Last summer, as I began to understand the thinking behind Doug's setup recommendations for our Cobra, I began to apply the same thinking to setups for GPL's cars.

And it worked! I found I was knocking typically at least a second off my previous best lap times, in both F1 and F2 cars, and I was smoother and more consistent as well. At the Nurburgring, I took over 7 seconds off my personal best with the Lotus, and made similar improvements with the other chassis.

Now you can download all the GPL setups I've done since September of last year, when I started to apply the same theories to GPL that we were using on our real-world Cobra.

What's Different About These Setups?

GPL's Tire Model

Thanks to an anonymous benefactor, I have been able to test suspension setups with a model of my real-world Cobra in GPL.

When I first began experimenting with this, I noticed that when its suspension bottomed the real-world Cobra behaved differently from the GPL Cobra. (To avoid breaking things, we've since fixed the real-world Cobra so its suspension doesn't bottom; details on my Cobra site.)

I found that while the GPL Cobra tended to snap sideways when the suspension bottomed, the real-world Cobra tended to track straight. Compressing the rear suspension to the point where the tire was the only spring did not cause the tail to step out, as it does in GPL.

I discussed this with a Papyrus engineer, and he responded that one area where NASCAR 4 improved on GPL's physics was that its tires modeled high vertical load behavior more realistically.

I pressed him for details, and he said that at high vertical loads, GPL's tire model produces a lower level of grip than a real-world tire would. N4 models this part of real-world tire behavior better.

I believe this tends to vindicate my previous use of higher ride heights to avoid bottoming the GPL car's suspension, and also explains why my Gen 2 setups feel more realistic to me than most GPL setups I've tried: they keep GPL's physics engine within the range where it models tire physics most accurately.

 

Sticky Slides

How can you tell whether the suspension is bottoming or not?

Sticky slides. When you go over a bump with a typical GPL setup, the outside rear wheel bottoms on its suspension and the tail snaps out.

Worse, it tends to stay out, and go even further, because GPL's tire model loses too much grip at high loads (see sidebar). With traditional setups, once the tail snaps out, unless you react very quickly, it's gone.

By contrast, when you keep the suspension from bottoming, the car does react to the bump, but in a smoother, much more controllable way.

The tail may step out, but once the bump is past, the car either catches itself or lets you catch it with a less abrupt correction.

You feel the bumps, but they don't make you spin unless you're really over the top.

The result is a more enjoyable, more realistic racing experience - especially with a good Force Feedback wheel.

How are these new setups different from my old setups? They are based on Doug Arnao's setup principles for real-world cars. Without going into great detail (and revealing all of Doug's secrets), Doug determines appropriate spring rates based on vehicle weight, weight distribution, and available wheel travel. Given these values, and the car's roll centers and center of gravity heights, he calculates appropriate damping rates and anti-roll bar rates.

I came up with some formulas to translate Doug's values for our Cobra to GPL's cars, which are lighter and somewhat more tail-heavy. Using the results of these formulas, I've gone to significantly higher spring rates, with damping rates and anti-roll bar values to match.

Like Doug's real-world Porsches and our Cobra, these new setups have the advantage of keeping the car stable as a handling platform in transients (e.g. the car doesn't bobble when you turn in) and over bumps. They have the additional virtue of keeping the tires out of the range of vertical loads in which GPL's tire model breaks down (see sidebar).

Although these setups use springs that are somewhat stiffer than were reportedly used on 1967 F1 cars, they are on the soft side for a modern non-downforce race car - and hugely softer than a race car with significant aero download.

They are also effectively softer than the vast majority of GPL setups available on the Web. How can that be? Read on.

When Stiffer is Softer

Because they prevent the suspension from bottoming on the bump rubbers, these setups actually have a much softer effective spring rate - the spring rate that the wheel actually sees in operation - than most setups used for GPL.

Most "soft" setups for GPL use low ride height (2.5" is typical) and springs so soft that as soon as the car rolls, squats, and/or hits a bump, the suspension deflects to the upper end of its range of travel and contacts the bump rubber.

Once the suspension contacts the bump rubber, the bump rubber becomes the spring. The effective wheel rate goes up astronomically; essentially the tire becomes the main spring. You've now got an ultra-stiff wheel rate - but generally at only one or two corners. The result is that grip at the affected corner(s) is lost, and the balance of the car is upset. GPL's tire model aggravates the situation (see sidebar).

With my Gen 2 setups, because the suspension isn't bottoming on the bump rubber, the effective wheel rate - which is all the car cares about - is linear throughout the suspension's operating range of travel. This yields much smoother, less twitchy handling, and better overall grip.

As a result, I don't need as much understeer to help me keep from spinning when I hit bumps, so I can set up the car with more neutral handling, which makes it faster in midcorner.

And because the springs are stiffer than my old setups, they allow me to run the chassis lower than before (2.5" in most cases) and thus take advantage of the reduced weight transfer - and increased cornering power - which results from the lower ride height.

The result is faster lap times and more realistic-feeling and more enjoyable handling.

Download the Setups

Download my Gen 2 setups below. You'll need Winzip to unzip them.

Gen 2 Setups for F1
Gen 2 Setups for F2
Gen 2 Setups for FD
All Above Gen 2 Setups
Gen 2B Lime Rock Setups for F2

Adapting a Setup to Another Track

I don't have setups for all tracks in the Gen 2 collections. With the number of tracks now available, that would be impossible. This is simply a collection of all setups I've created since September 2001, when I started applying Doug's principles to GPL setups.

However, I use essentially the same setup everywhere, except at tracks which generate very high vertical G's (such as the Nurburgring). So if you want to run at a track for which I haven't supplied a setup, just do what I do: combine my latest suspension settings with some existing gearing for that track.

Note: I make sure the Gen 2 setup I "borrow" is for a similar type of track (i.e. I don't use a Nurburgring setup elsewhere unless it's a track that generates very high vertical loads). Silverstone, Snetterton or Zandvoort setups transfer well to almost anywhere else.

Here's what I do when I go to a new track, or one where I don't have a Gen 2 setup yet:

  1. I run Windows Find and find the latest setup I've got for the chassis I'm going to run (F2 Ferrari, F1 Eagle, whatever).
  2. I copy this setup to the setup folder for the track I'm going to run.
  3. I rename this setup to a name that identifies it as a new setup for this track.
  4. I run GRE and open this setup.
  5. Next, using GRE, I copy the gearbox from my last (old) setup for the same chassis at this track into the newly copied setup, and save the new setup.

Here's an example. Say I'm going to run the F2 Ferrari at Rouen. I haven't got a Gen 2 setup for the F2 Ferrari at Rouen. So:

  1. Using Windows Find, I look under gpl\players\Hine_F2__Alison for my latest F2 Ferrari setup. I find it's Ali Fer F2 Mosp x1e.
  2. In Windows Explorer, I copy this setup to gpl\players\Hine_F2__Alison\setups\Rouen.
  3. Still in Windows Explorer, I see that my old setup for Rouen was Ali Fer F2 Rouen x2d. I rename the new setup to Ali Fer F2 Rouen x3.
  4. I run GRE and open Ali Fer F2 Rouen x3.
  5. In GRE, I copy the gearbox for Ali Fer F2 Rouen x2d to Ali Fer F2 Rouen x3 and save the latter.

Now I've got my old gearing for this track (which was probably pretty good already) and my latest suspension settings.

Note: since the F2 cars get close to their top speeds at most F1 tracks, often no gearing changes are needed. However, the F1 cars usually will need appropriate gearing for each track.

Now I go on the track and drive a few laps. If it feels good, great.

If it understeers (pushes) a bit too much, I soften the front anti-roll bar by one or two clicks and stiffen the rear by the same amount. If it oversteers (it's too loose), I do the opposite.

I also may have to adjust the brake balance, but rarely by more than one or two clicks. If it's unstable under braking, raise the front braking percentage; if the front brakes are locking, lower the front braking percentage.

Also you might want to look at the tire temps. You may want to tweak up the front camber if the outside edge is running hotter than the inside, and the car is understeering. Same at the rear if you've got oversteer.

But don't go too wild on camber or the car will become unstable under braking. I find that with these setups, .5 or .75 degrees of camber is plenty.

That's it. Enjoy!

Update: F2 Setups for Lime Rock

November 2002 - Influenced by my ongoing real-world track experiences with my Cobra, I've continued to evolve my setups. I've added some F2 setups for Lime Rock to my Gen 2 Setup setups.

These new setups are the most balanced setups I've ever released. They are so different from my earlier setup that they should probably be called "Gen 2B".

This new approach to balancing the car goes back to the development we've been doing on the Cobra this year (2002). Nate kept complaining of understeer. I wasn't so bothered by it, but that's because I was accustomed to trailing off the brakes after turn-in, using the weight transfer to get the tail to step out before I got on the power.

Nate, however, doesn't trail brake, and he needed the car to turn in after he got off the brakes, and then not understeer when he got on the power. But the Cobra had serious understeer with the power on.

We did a lot of development on the suspension this summer and fall (read more about the development process on the Events pages on my Cobra site), and finally in October got the car balanced well enough to make Nate happy. I liked it better too, and since Nate was by now consistently quicker than me, I figured I'd better try to change my style to be more like his.

When I got home from the October event at Loudon, I did some laps in the GPL Ferrari with my latest setup. I was surprised at how much the car understeered compared to the now-neutral real-world Cobra! I thought, Wow! So this is why all the really fast drivers complain that my setups have too much understeer!

I started tweaking my Gen 2 setups to get better balance. I kept the spring rates and ride heights the same, but fiddled with the differential locking, anti-roll bars, and rear shocks, trying to get the car to turn in well, without excessive midcorner understeer, and respond to throttle with smoothly progressive power oversteer - and yet still be forgiving if I carried a bit too much braking through turn-in, or otherwise got the tail out a little too far.

I have had to change my driving style as well as my setup philosophy. I used to rush up to the corners, toss the car in and trail off the brakes to make it turn. I hardly ever had to put in opposite lock; I was always just coping with understeer. If the car did get sideways, I had dialed in so much trailing throttle understeer that all I had to do was lift out of the throttle a bit and wait, and the car would scrub off speed and hook up again. It was a very safe and forgiving way to set up the car and to drive, but it isn't the fastest way.

Now I've had to learn to cope with oversteer. My new setups, like many setups used by aliens and other fast GPLers, tend to oversteer, especially under power. My hands are constantly moving on the wheel, turning in, then back out to balance the car as I pick up the throttle, then to opposite lock on corner exit. A more deft, delicate touch on turn-in is necessary, but it pays off with better midcorner speed.

Having really good force feedback helps, because it lets me know when the car is transitioning to oversteer by the pressure in the wheel; the caster tends to pull the wheel toward the outside of the turn, into the slide. I can recognize a slide by feel before I recognize it visually, so I can live with more oversteer than I used to.

The fundamental philosophy of the Gen 2 setups (higher spring rates to keep the car from bottoming at low ride heights) keeps the car stable enough that it doesn't do anything really suddenly, which helps me deal with the oversteer I do get. It reacts to bumps, but in a realistic way, and that's much easier for me to cope with than the sudden snap oversteer that occurs with more conventional low-rider setups.

The result is the best setups I've ever done for GPL. I find these setups to make GPL's cars feel more like a real race car (such as my real-world Cobra) than any other setup yet. They also make the cars more rewarding to drive, because with practice I have found myself balancing the car on the fine edge of traction with the throttle, front and rear tires running just about the same slip angles. It's a delicious feeling! In fact, I think it's the best sensation of all in motor racing.

These setups are fast, too. Joe Tyler tested my Silverstone F1 Ferrari setup and did a 1:26.73. He made a few tweaks and managed a 1:26.11, two hundredths off his personal best!

My latest F2 setups, for Lime Rock, have been influenced by Joe's tweaks to my F1 Ferrari. Thanks, Joe!

Anyway, give these setups (so far I've done F2 Ferrari, Lotus, and Eagle for Lime Rock) a try:

Gen 2B Lime Rock Setups for F2

You can also download some Lime Rock replays of laps done with these setups. See the F2 grid on my Replays page.