not at all am I saying ignore the math. but be sure to verify using real world testing, and I don't mean test using the "butt dyno." collect data, lap times, temps, traqmate, and see if your math is really making a positive change.

 

^^ True.dat. The "numbers" don't take into account things like effects of tire spring rate, roll centers, chassis stiffness, bushing deflection, sticktion, etc.. Im just advocating using the math to get in the ball park at least and then deviate but deviate with an idea of how it should impact the balance or grip of an end.

On a side note, I don't believe in small changes unless you are certain of the result. For instance, small shock changes are usually meaningless and impossible to feel. But if you run 10k/10k springs and know that balance works but need just a little more, then 11k/11k might just work out.

 

yeah i agree whole heartedly with your approach. cars in storage right now at my parents house until next track outing next month. i will try to get it weighed when early next month so i can start trying some math.

in the mean time i might pick up a set of 11k springs...however im concerned it might be a huge change to the front and screw me over. then the question will be if i go back to stock bars...

 

why would you go back to stock bars?

 

^ . I don't know what Grice runs, but I have pretty big bars front rear.

 

I find it interesting that Ralliart suggests a large front bar for tarmac use.

 

I thought that eventually it became better to control the roll with springs and less bar? something about it can help keep tires in contact with the ground? but again its very possible i completely misunderstood all of that.

 

It's been said to "follow the math" but what math specifically?

"Conventional wisdom" such as Milliken suggest to pick spring rate based on natural frequency and then use the sways and geometry to correct the handling balance. The target by his metrics is a frequency of ~2.0 Hz in the front on a non-aero car with the back 0.1 to 0.2Hz higher for bump stability.

Math could also lead you down the road of roll stiffness calculations. In this realm, you can completely balance weight transfer without sways at all. The natural frequencies will be all screwed up though.

Finally, math can also take you down the road of forced based reaction calculations. The most accurate as it accounts for geometry changes durning Corning loads but it's also very time consuming to build up correct models and simulations.


In the world of conventional wisdom, square spring rates front and rear are the correct method. 10kg f/r puts you around 2.4/2.6Hz on a 3000 lb 60% front biased Evo.

 

Also it is important to realize what kind of racing is being done when you evaluate other people's advise.

Most guys running high rates in the rear are autocrossing. Generally speaking, the Evo is a pig at low speeds. High rear rates make the car pretty twitchy in the rear and throttle/brakes upset traction in the back before the front this way. It makes the Evo so you can throttle steer and trail brake it through a tight corner.

"Twitchy" probably isn't how you want to describe the ideal setup on a road course car that regularly sees corners at 80+mph.

 

you're absolutely right, i like the balance of my car now as the rear is never an issue.

gonna try to run some numbers this week when i have some time and see what i come up with for a basically stock weight car.

 

Yes, these are the two things I follow. We run in the 2.6-2.7 front and 2.8-2.9hz rear range with an additional 30% front stiffness and 35-40% due to sway bars. Is this the best way? I dunno, but when I try other things I get slower or end up with a twitchy car that's harder to drive. It could probably use more grip steady state up front but thats an Evo issue in general with the 65% front weight we carry.

For an interesting bandaid to what I though was going to be torrential rain this Sunday, I ran a 26mm front bar at full soft (around 70% original stiffness with cusco brackets) and no rear bar on 255 ZII's since I don't have a set of rain race tires.

Conventional wisdom says this will push like a pig but surprisingly it didnt do too bad. I didn't have mid corner rotation or any lift rotation if I needed for tight sections but front grip was actually pretty decent. Still had plenty of throttle rotation and I might say it was faster through a slalom than I expected.

 

Yeah, the more of the math approach I've taken the more I've realized math alone only gets you a subset of possible solutions to any given problem.

For example, math can give you a line of sway bar vs spring stiffness to reach a desired roll stiffness. The problem then becomes, which ratio is actually the fastest.

Then you look at how changes to roll height changes this subset...

Experience is probably more important than math here. That's hard for me to say but I think it's the truth. I've been messing with this stuff on another car anyway and I still have yet to really sort the car out. I went straight math for the baseline and it didn't work well at all...

 

These links from Works seem to do a decent job explaining the math
http://www.worksevo.com/technical.php

 

you follow the math to happy tires.

on an evo that usually means managing camber control with high roll stiffness, particularly when it has a short sidewall tire on it.

to me, square spring rates are compensating for inadequate rear camber.

 

I would guess due to the imperfect roads on your average tarmac rally, they might want more of their roll resistance from the bars and less from springs?