Spring rate theory and application regarding our abnormal chassis.

[BenFenner]

This is a thread for us to discuss spring rates in general. Theory, practical application, and everything else is welcome. This thread is here to consolidate some conversations that have been on-going in other threads in the past.

Here are other threads/links on the topic:
http://www.bimmerforums.com/forum/sh...e-spring-rates
http://e30m3project.com/e30m3perform...e/eff_rate.htm


Here are threads partially derailed by spring rate discussion if you want to read through them:
TC Kline D/A coilover refresh - or - fixing suspension noises

A quote from this thread to get us started:

That's the thing. TCK calls for rates like that [400 lbs/in front, 400 lbs/in rear], and every one seems to stick close to it. Going by spring rates doesn't really tell the story. It's the wheel rate, what happens where the tire meets the ground, that matters. The wheel rate is the spring rate times the motion ratio of the suspension squared. Thus, kibitzing off an E30 racing forum post, the front would be spring rate x 0.88, and the rear would be spring rate x 0.45.
So in theory with our near 50/50 cars, you would want springs that are twice as stiff in the rear just to keep an equal front/rear wheel rate. While dead equal is a bad thing in practice, I can't see deviating as much as seems to be the normal choice for these cars. What the hell am I missing?

What are the stock spring rates? That would be a good place to start IMO.

[Msilvia]

This thread has great info/calculations as far as creating a neutral balance for our cars(E30+E36 frankensteins):

http://forums.bimmerforums.com/forum...s-for-drifting

[BenFenner]

Well, the info in that thread corresponds with what rf900rkw (and I) have been saying about needing a good bit more spring in the rear simply to approach an equal amount of "wheel rate". It also makes a good point about proper coil-over setups in the rear. That is obviously NOT part of this discussion. Changing the location of the spring changes things completely (and makes things more "normal" for spring rate needs). We can stay away from that in this thread (for now?).

I haven't done the math yet (I'm one who likes to do my own math) but I'm kind of skeptical about having to square the motion ratio before multiplying by the spring rate to get the effective wheel rate. That doesn't seem right to me. The mechanical advantage from a [class 2] lever and how it multiplies force shouldn't involve squaring the mechanical advantage I don't think. It should work linearly on the spring, i.e. no squaring needed. Time to dig back into rudimentary Newtonian physics.

[The ZED]

I thought convention yielded a 100lb bias between front and rear spring rates? 400/500 450/550 etc. I know on my car it felt much better with the stiffest rear spring I put in there (550lb).

[BenFenner]

Anyone know the stock rates?

(I've already asked that a mod correct the spelling in the thread title.)

[The ZED]

Found it, Found it:
https://spreadsheets.google.com/pub?...3I3mP8cOvc5vZQ

[BenFenner]

So basically 80F/200R for some variant of Z3 chassis (averaging the two front spring options together). And they seem to be slightly progressive. So they are really 80-88F/200-225R. I'm not sure if that is on purpose or if they are just static rate springs of poor consistency.
That makes me even more confident in rf900rkw's assessment that rates should be about doubled for the rear (theory only). I know he was hesitant to go that far outside the norm for his own setup though. I'm thinking a 350F/700R setup wouldn't be nuts. Or maybe a 275F/550R for those wanting more compliance.

[JoshS]

There is so much more to consider.

Swaybars basically add spring rate in cornering but not in bump. So you can sometimes get the desired additional rate effect with bigger swaybars. This has other penalties of course, and at least one suspension design guru I've talked to recommends getting the spring rates right and then getting "balance" by adding ONE swaybar.

Don't forget that any part that moves is basically part of the suspension and contributes to the effective wheel rate. If your chassis deflects at all, stores energy, and then returns to its original shape, then it is acting as a spring. Same as the tire sidewalls. Mathematical tuning of springs and shocks, when the chassis is flexy and the sidewalls are soft, is just fooling yourself. And if you go too stiff on the springs/shocks/swaybars with those assumptions, then the chassis and the tire sidewalls is forced to contribute even more to the equation. Bottom line is that the mathematical models don't apply all that well to street cars (especially flexy convertibles) for this reason, and anecdotal evidence on specific cars is probably just as valuable.

[BenFenner]

We know there is so much more to consider Josh. We're just trying to get a base-line to start from I think. And we are talking in large, sweeping generalizations. Right now I don't think we want to get into the serious nitty gritty.

In general, it would seem a 400F/400R recommendation for spring rates is out of left field. Would you not agree? Assuming stock or relatively mild performance components elsewhere, such as bushings, swaybars, tires. Same question for the suggested norm of 400F/500R or 450F/550R. Is that not a little too far off the reservation?

Doesn't it seem odd that you would more-than quadruple the front rate, and only double the rear? Especially considering the mechanical advantage built into the rear? Do we think the Z3 chassis is that poorly balanced from the factory that it needs so much more front rate?

[FinnNiks]

I went with a 400/500 rate for my TCK S/A's. The TCK rep I talked to when I was ordering recommended it.

[fmcfad01]

I have a set of 600# rear barrel springs if anyone wants to borrow them for science. I bought them years ago and never installed them.

[rf900rkw]

I have a set of 600# rear barrel springs if anyone wants to borrow them for science. I bought them years ago and never installed them.

I have been thinking of going to 600# in the back, but really need to revalve the shocks for that much.


Ben -- It is squared. The motion ratio applies to both the force and the distance. Take a simple 2:1 ratio. In order to apply 100 lb/in at the wheel, the spring needs to have 200 lbs of force in 1/2 and inch; 400 lb/in total.


And I will repeat in this thread... I am not saying that the TCK-ish rates are wrong. As Josh said, there are many factors that add and conflict with each other. Often you have to make a compromise on one section to prop up a weakness elsewhere. Years ago I was doing SRF, and developed a very unorthydox suspension setup with the roll couple distribution jacked heavily to the front, much like the TCK setup does. It flew in the face of what everyone else was doing, and netted us multiple division titles and a Runoffs podium in the only appearance. And my avatar.

So I'm looking for technically why the TCK way is better, if it is. What is being propped up by so much front spring? "Cause Pete says so" tain't the answer we seek.

[BenFenner]

Ben -- It is squared. The motion ratio applies to both the force and the distance. Take a simple 2:1 ratio. In order to apply 100 lb/in at the wheel, the spring needs to have 200 lbs of force in 1/2 and inch; 400 lb/in total.

I started out doubting this claim. I started writing out a thought experiment to show my confusion. In the process I eventually saw the issue and it makes sense to me now. Here's the resulting thought experiment for others.


Take 100 lbs. and put it on a normal (class 1) lever with a 1:1 motion ratio acting on a 100 lb/in spring and that spring will experience 100 lbs. of force, and compress 1 inch. The end with the weight represents our wheel travel distance, and in this case it travels 1 inch as well. This is our control. 100 pounds of force resulting in 1 inch of wheel travel. This is a figure of work (force × distance). In this example we've done 100 in·lbs of work.

Take 100 lbs. and put it on a lever with a 2:1 motion ratio acting on a 100 lb/in spring and that spring will experience 200 lbs. of force, and compress 2 inches. Right? The other end of the lever (where the wheel is) will travel 4 inches. To match our control, we want the wheel to move 1 inch total, correct? In this example we've done 400 in·lbs of work.

Take 100 lbs. on that same 2:1 lever acting on a 200 lb/in spring and that spring will experience the same 200 lbs. of force as before, and compress 1 inch. Right? The other end of the lever (where the wheel is) will travel 2 inches. This still doesn't match our control. In this example we've done 200 in·lbs of work.

Take 100 lbs. on the 2:1 lever acting on a 400 lb/in spring and that spring will experience the same 200 lbs. of force as before, and compress ½ an inch. Right? The other end of the lever (where the wheel is) will travel 1 inch. This matches the control. 100 pounds of force resulting in 1 inch of wheel travel. In this example we've done the same 100 in·lbs of work as in our control.

[Arkangel]

Ground control recommended me to use the same rates on my Z that they do on their E30 - 440(450?)/660. Ride is actually alright.

[albrazzi]

So if stock rate is 88/200 (82/134 wheel rate) and all traditional wisdom is that higher (relative) spring rate, and Swaybar rate Front, as evidenced by higher rate Delta Front to rear and softening or removing the rear bar altogether depending on spring rate is what improves handling on our strut cars, then any of these debated combinations will do that and its only a matter of how many combinations you are able to feel to dial in the personnel choice. Springs get expensive for some people.
It would seem to me anything from 400/400 (376/268 wheel rate)to 450/550 (423/367 wheel rate)etc. will stiffen the front end (relative) to stock rates and make for better handling. But Im repeating myself. You would have to go more like 440/1000 (413/670 wheel rate) to accomplish the same spring bias as OEM and I'm not suggesting that. Ultimately some street driveability is compromised on a true Track rate setup.

I hope this adds to the discussion since as I see it the purpose is to improve Handling and not to simply raise equally the rates and maintain OEM understeer.

[BMWBergy]

"Cause Pete says so" tain't the answer we seek.

Because TC said so? I know, that wasn't right!

[rf900rkw]

Al -- wheel rate is spring rate times motion ratio squared. That would make the stock 78/90 and so on. This, however, doesn't change you point of the front-rear ratio being what it is.

There was a thread a few years back where we got some of the diehard autoxers talking springrates and one, Car54 if I remember, admitted to rates approaching 1000 lb/in. It may have been 800, I just remember being surprised at the resulting frequency/rate.

[JoshS]

Springs get expensive for some people.

But cheaper than sway bars, and WAY cheaper than good shocks/struts!

[BMWBergy]

Found it, Found it:
https://spreadsheets.google.com/pub?...3I3mP8cOvc5vZQ

Wow. Just saw the spring rates for the Dinan E36 M3 springs... 100# Front / 425# Rear...

[The ZED]

But the problem is the ride height difference. We are asking it to perform in a shorter operating area then stock, but how much??

[albrazzi]

But cheaper than sway bars, and WAY cheaper than good shocks/struts!

True but after buying all that stuff and wondering what difference a new spring would make the money always goes somewhere, else for me anyway. Currently I am on AST 4100s RD sways and have already put the OEM bar back on the rear and 450/550 (their middle choice) I am too stiff for highway cruising and rough roads but just right for the Track even the higher choice would be better for most Tracks and worse for the street. But thats not the debate I guess.

[BenFenner]

The turbo e30 we built went from stock suspension to progressive (so I'm told) IE "stage3" race springs controlled by Bilstein "sport" dampers. The ride height dropped significantly, the roll control was great, and the ride quality was almost indistinguishable from stock. I wonder what their rates are.

I have a set of 600# rear barrel springs if anyone wants to borrow them for science. I bought them years ago and never installed them.

If I can find out what my front rates are on the coupe those rear springs might make for a good test on my car. Assuming the Konis in the rear are up to the task once adjusted for the increased rate. If I have around 300 lb/in springs up front this could be interesting. Although right now I have a wacky base-line of E36 springs all around (I think). Maybe my car wouldn't be the best comparison right now.

[z3papa]

True but after buying all that stuff and wondering what difference a new spring would make the money always goes somewhere, else for me anyway. Currently I am on AST 4100s RD sways and have already put the OEM bar back on the rear and 450/550 (their middle choice) I am too stiff for highway cruising and rough roads but just right for the Track even the higher choice would be better for most Tracks and worse for the street. But thats not the debate I guess.

Something is off on your setup. I've run every spring rate front (from 400-700) and rear (500 to 800) with varying settings on bars, no bar rear, rebound/compression settings and tire pressures in search of the best different setups for asphalt and concrete in dry and wet. Much of the testing started with the discussion of motion rates such as outlined here, considerations for additional bar rate (Josh's comments) and chassis flex. For the record, you can buy used GC, Hyperco, or Swift springs in pairs for $60-70 shipped. I have friends with libraries of springs 5"-6" in length. I know some believe in the 100# split, and others discount the adjustment for rear chassis motion rates. Since I race on street tires, I tend to have a "softer" rear to get more rear end squat grip while giving me a car with crisp transitions. My car is pretty comfortable on the highway and I still take it on trips to cities without concern while running right now on 575/700 with H&R bar on soft up front and rear OE bar attached.

[rf900rkw]

Just to toss out a few comments.

Bars do not add spring rate, they *move* spring force to the outside wheel by stealing it from the inside wheel Lots of spring and no bar is not the same as a little bit of spring and a lot of bar. Both ways can end up with similar weight transfer characteristics, and end up with the same cornering potential, but how the dynamics of how it gets there is very different. This gets to be very important when forces change, such as adding a longitudinal vector to the lateral load, aka powering out of the corner.

The adjustment on Koni SA, and the rebound side of the DA, is for the low speed rebound valve only. This part of the shock dynamics comes into play during weight transfer, and can be used as a tool to trim the cars handling. Changes to ride comfort (high speed compression valving) and spring control (high speed rebound valving) require a trip back to Koni.

[BenFenner]

Trying to keep things spring related here.
If I wanted to try some much stiffer springs in the rear, I would obviously want to get a damper with appropriately stronger rebound damping. rf900rkw, you're saying that needs to be stronger high speed rebound specifically, correct? That makes sense.
And the adjustment for the single adjustable Koni is for compression anyway? And low-speed at that?

So with my single adjustable Konis, I'm not compensating for spring rate with those adjustments... I'm controlling low-speed compression. Which is basically a way of adding yaw/roll control while trying to maintain ride quality. That is not how I prefer to tune a chassis, that's for sure.

It might be time for me to get a proper setup on my car, versus living with what the previous owner tossed on there.