[stretch]

I measured all the various bits of my car, somewhat inaccurately, with a tape measure. My calculated roll center height agrees with Whiteline's measurement. To my surprise, the STI's roll center is not as low and the camber curves not as bad as I originally thought.

Here is a diagram showing what the geometry looks like stock (gray lines) versus 1" lowered (colored lines):


In the chart to the right, the numbers under "dive" would be the same numbers you'd get from lowering the car.

So, you can see in the numbers that at 1" of dive, we get 0.59 degrees of camber gain when body roll is not present. At 2" of dive, we get -1.05 of camber gain, which is 0.44 degrees more than at 1" of dive. Therefore, if the car was lowered 1" to start with, it'd only get that 0.44 degrees of negative camber in its first inch of compression.

You can see then that the camber curve gets worse as you lower the car, but it starts out pretty decent.

Going left or right in the table will show you individual wheel camber change during body roll, but keep in mind that this is body roll that the suspension sees. Most stiff, modified Subaru's will have a significant portion of their body roll coming from tire deflection, which means that while the car's body is tilted 4 degrees total, the suspension may only be tilted 2 degrees relative to the tires. Thus, the suspension only sees 2 degrees of roll.

Notice that for a stock car to roll one degree, it'd gain -.64 degrees of camber. However, if the car is rolling one degree, we get 1 -.64 degrees, or positive 0.36 degrees of camber total. That means we actually have positive camber in the turn: yuck. Camber curves are never as fast as body roll, which is why we need to run static camber.

[stretch]

Here's how the camber curve changes with camber plates moving your strut mounting point one inch inward (maxed out for most products):


Notice there's quite a bit more dynamic camber when doing this! The roll center height goes up a bit, but not much.

[stretch]

With a Whiteline Roll Center Adjuster kit:


Notice the roll center does go up by 0.8 inches over stock (or an inch for a lowered car), but the camber curve does not quicken as much as with camber plates set fully inward.

[stretch]

Why you want your tires perpendicular to the ground, all the time:


Thus it is the suspension geometry's task of keeping the tire as vertical as possible over the widest variety of conditions. Unfortunately, that's rather difficult to do, as seen above. The suspension geometry cannot change camber as quickly as needed to compensate for body roll.

Why camber improves turn-in:


What this shows is that a tire generates lateral thrust (steering) just from having camber. Thus, camber improves steering response because the tires always want to turn. The moment you put more weight on a tire, the car is going to turn because the tire was already trying to do that.

However, this is also why stability goes downhill with lots of camber. Your tires will continually fight one another for grip; each trying to turn inwards while you fight to keep the car straight. It's going to feel almost like running lots of toe-in up front. However, since your tires are fighting one another instead of working together, you will lose straight-line grip. This is seen in the previous graph: your grip under braking will drop rather considerably with lots of camber.

[stretch]

Roll center heights.

Edit: discussion here, instead: Roll Centers | Build A Faster Car

[fullerton]

What about things that modify caster (ALK & Camber/Caster plates)? Will that do anything since the strut won't move up the same amount (as compared to the ground) per inch of compression?

[mgsm666]

*subscribed* I LOVE this forum...you guys are awesome, and never fail to amaze me. Thanks for taking the time to do this! Very helpful

~Jesse

[Greg]

Originally Posted by stretch

This spot reserved for showing the effect of maxing your camber plates.

When you run the numbers for adding negative camber using camber plates (and increasing your KAI), can you also include dynamic camber changes caused by not increasing caster at the same time? For more info, see this post and follow the links.

[stretch]

Originally Posted by Greg

When you run the numbers for adding negative camber using camber plates (and increasing your KAI), can you also include dynamic camber changes caused by not increasing caster at the same time? For more info, see this post and follow the links.

Sorry, the software I'm using is very basic and does not do that. The more comprehensive software costs money, unfortunately.

[javid]

Nice work, looks pretty good so far. What are you using for modeling?

Is your car at stock hieght? Hub to fender distance please?

[stretch]

Originally Posted by javid

Nice work, looks pretty good so far. What are you using for modeling?

"Roll Center Calculator v2.0"

I only have a free demo which doesn't let you manipulate the values... unless you open up a saved file and edit that manually! (Luckily, the file is saved in clear text, easy to understand.)

Is your car at stock hieght? Hub to fender distance please?

Yes, when I took these measurements I was at 14.75.

[scoobydo]

Stretch:
Would you mind posting all of your geometry measurements/input? I would like to tweak the numbers a bit myself. No sense in starting from zero.

Thanks in advance...

[stretch]

Sadly, I no longer have the software where those numbers were stored. I wish I did! It didn't take too long to take the measurements, though.

[scoobydo]

Well, are you happy with the results? What would you do differently? I am running a stock 06 and would like to get rid of the understeer. Not very noticeable till your running an autocross, or track event and then it is very frustrating.

[MGizzle]

If you are runing stock getting more camber in the front will kill your understeer, if this is your only concern. I actually started loosing the tail too much when I went to camber bolts and -2.1deg and went back to stock bolts and -0.9deg in the front. Camber bolts suck, slip too much.

Anyhow, I would recommend driving your car a lot and "learning" the understeer. You will find once you are really familiar with your car that it is not that bad if you handle it right.

Finally, even when you do get your settings down the STi seems to offer plenty more room for improvement. I actually requested pictures from friends of my car rolling at the track (my ride is stock) and I have 1-2 shoots where my entire inside (front and rear) is almost lifted off if not already up due to the enormous roll.

Stretch, any idea if scrub (not scrub radius) will increase on the McPherson layout if we increase roll center?? I know it does on a double wishbone but not sure about our geometry. Hmmm...got to look into that one.