Nohachi's suspension ramblings thread

[NoHachi]

How much is too much?
So after looking at all these frequencies etc you kinda start to wonder..who to believe: the JDM companies with years of experience or some simple number that corresponds to some frequency..what's up with that? You would be wise to be more then a little sceptical.. So I started thinking, how much really is too much? Yeah AE86 spring rates tend to be high..but their suspension is pretty archaic so you need plenty of roll resistance to compensate. Maybe that's what's going on? Then something popped into my head. What exactly is doing the suspending here?

Twister: a tale of chassis flex


Years ago, in my younger years I took my chariot of that time, a little civic, to the nurburgring. I threw the poor old thing into a particular corner hard, desperately trying to stay ahead of a friend in his M3 (I cheated by learning the track layout in GPL). My passenger made a comment that the view was really quite beautifull. Especially since he could see so much of it through the gap between the passenger door and the roof. Que!? Well, the little civic's chassis was twisting quite severely you see. Enough to open a gap of 2 to 3 cm's in high G corners.

So what influence does the 80's econobox chassis have on our choice of springrates?

Simple
If you think about it its quite simple. Let the tire push one corner of the car up and either the spring or the chassis will flex. Exactly how much each of them flexes depends on their relative stiffnesses. An infinately stiff spring in a weak chassis will see the chassis do most of the flexing. A stiff chassis with weaker springs will have the springs do most of the work.

What about the *86?
So what about the old 86? There are some simple calculations we could do. With a bit of math you can translate the chassis stiffness to an equivalent stiffness at the wheel. Its a simple case of calculating how much a twist of a degree at the chassis ends up being at a suspension pickup point. If you want to know more on how to model this yourself, take a look here:
http://www.theoryinpracticeengineering.c...20fsae.pdf

You can do a full car model if you want, but I'll stick to a simple and indicative single corner model for now. There is a problem: what is the torsional stiffness of a hachiroku? I have never seen a number, so if anyone has found some somewhere please post it up! Without a definate number all we can do is guestimate a range for the number. Here are some references:

Mazda NA MX5 (1990): 6000Nm/deg
Lotus Elise S2 Exige (2004): 10,500 Nm/deg
BMW E46 Sedan (w/o folding seats) 18,000 Nm/deg
BMW E46 Sedan (w/folding seats) 13,000 Nm/deg
BMW E46 Wagon (w/folding seats) 14,000 Nm/deg
BMW E46 Coupe (w/folding seats) 12,500 Nm/deg
BMW E46 Convertible 10,500 Nm/deg
(find a small list here: http://www.germancarforum.com/test-data/...idity.html
Note that chassis rigidity has really improved in recent years. Part of the reason why modern cars hold the road quite well).

So where do we estimate the AE86? It must be stiffer then an old MX5..but stiffer then an Elise? Hmz..they are convertibles no? Looking at the BMW 3 series we see that sedans are quite stiff, but isn't the E46 a more modern design? (answer: it is).
At the very least we have a range to work with: The 86 has a torsional stiffness that is probably better then a miata's, but less then a E46 sedan's. So roughly between 6000Nm/deg and 14000Nm/deg.

What to aim for?
At first I though that maybe we would get chassis springrates less then those of the actual springs.. I'm glad to say that isn't the case. I remember from my old FSAE days that we used to aim for a chassis that's about 10 times stiffer then the spring. So for each 10cm's of spring movement we would get 1 cm of chassis flex. I searched around for references but didn't spend to much time. One I found was the following:

Quote:To keep the chassis from doing the suspension's job, Mazdaspeed's engineers designed the wheel rate to be six to eight times softer than the chassis stiffness. In other words, the chassis deflects one inch for every six to eight inches of wheel travel. To calculate wheel rate, divide the chassis' torsional stiffness by the target ratio-7.5 in this case. Finding the appropriate spring rate is then a matter of knowing the suspension's motion ratio and doing the math.

From: http://www.modified.com/features/0601_sc...cross.html

Pretty simple eh! We can do that!

The results
Included (hopefully) in the attachment is an overview of the results. I did the calcs for the front since the front carries the highest wheelrates. Entries in green are combinations of springs and chassis where the chassis is more then 7.5 times as stiff. Red entries depict combinations where the ratio is less then 7.5. Pick your own estimate of the chassis' stiffness, plug in your springrates and away you go!



My car carries ~6kg/mm front springs..hmz guess if I want to have the chassis work optimally I will have to start thinking about some serious bracing. Some spotwelding or a cage definately seems in order! Maybe thats why those modifications are so popular in Japan!

To summarize:
*To run 5kg/mm springs on an AE86 you need ideally to stiffen your chassis through seem welding.
*To run 6kg/mm or more you really need a cage.