THSCC Forums

As several people have commented, the springs on the Duff car are very soft and the sway bars are large. When I originally planned it, I had a lot of other problems. The drivers were prone to black flags and tearing up the car, the team strategy was mostly nonexistent (and the strategery that existed was not good), and the powertrain liked to blow up a lot. At some point I figured I'd put a stock ITR spring/sway-bar combination on it and that would work okay while we sorted out the other stuff. It did work okay, but 250lb/in on all 4 corners is probably robbing us of some grip.

Now that the drivers, team, and powertrain are fixed, it's time for some suspension, which in this case probably means springs, because my budget is too low for a set of Motons, but $200 for a new set of custom rate springs is feasible at this point. The other motivation for this is that we're mostly likely getting bumped up a class for the upcoming CMP race, so maybe it's time to get faster.

Looking at what the cool kids are rocking, I see racing spring sets in the range of 450-550 front and 350-450 rear, which is in the range of what some guy at Ground Control was talking about when I talked to him. He also thought that my Biltein B6 OEM-replacement shocks would handle those spring rates.

Having fun with math and suspension design theory, the current setup gives me around 2Hz in the rear and a few tenths less in the front, which I'd expect from a production street suspension. The sway bars beef that up considerably under hard cornering, but I don't feel like doing enough math to figure out how much. Conventional wisdom (meaning all of the stuff that I could find to read combined with stuff that strangers on the Internet posted) on non-aero cars says 2-3 Hz should be my target, with the rear around 10%-20% higher than the front, so 525lb/in up front for 2.2Hz and 425lb/in in the back for 2.5Hz should work better and seems in line with what others are doing. The sway bars would be discarded.

I'd be interested in hearing others thoughts on this, since this is my first attempt at doing suspension math or coming up with a suspension for a dedicated race car. All measuring/weighing/math was done by me, so the numbers should be mostly correct.

_________________
"Every ship is a minesweeper, once."

Brace for long winded, technical response!


I approached this from a slightly different angle – keeping in mind that spring selection is a compromise among competing priorities. You need to know the desired ride height, the available suspension travel, the car weight (ideally the corner weights), the relationship between wheelrate and springrate (determined by suspension geometry) and finally the road characteristics you are likely to encounter.

Using my car as an example, we'll start with the front:
Let’s say we’re going to lower the car to the point where we have 4” of suspension travel before hitting the bumpstops. When last on the scales I had LF=844lbs, RF= 848lbs. For simplicity, let’s say 850lbs each. I haven’t weighed the wheels, tires, knuckles, control arms, strut, spring and CV shaft, but I guessing there’s about 150lbs of unsprung weight at each corner - leaving 700lbs of sprung. Let’s assume we are going to be driving on a fairly smooth track so we design for a maximum 1.5g bump. Based on these assumptions, we would be absorbing a 1050lb force over 4”, therefore requiring a minimum wheelrate of 262.5lb/in.

Of course wheel rate and spring rate are not the same, but they a related by a constant that is a function of the control arm linkage ratios and the instant center of the front suspension.

Time for some formulae:

WR=C*SR
or
WR=(MR)^2*SR*ACR
where
WR = Wheel rate
MR = Motion Ratio
SR = Spring rate
ACF = Angle correction factor or sin(θ); θ = the angle between the LCA and the strut

Strictly speaking: MR=SQRT[(a/b)^2 * (c/d)^2]
where
a = LCA inner pivot point to the point where the spring attaches to the LCA
b = Length of the LCA from the inner pivot point to the ball joint
c = LCA ball joint to the instant center
d = Center of the tire contact patch to the instant center

The instant center is the intersection of two imaginary lines: one formed by the angle of the UCA and the other formed by the angle of the LCA. This is where you can have some variability. Also, the instant center changes with ride height.

On many cars you can simplify: MR=a/b and the answers should be very similar. The dominant term is a/b. Being off a couple of degrees and even a couple of inches in the c/d measurements makes little difference, but a 1/4" error in the a/b measurements makes a big difference.

On the front of my car I found
a=9.25"
b=12.625"
c=163"
d=167"
θ = 75deg. (remember this changes with ride height also)

After some number crunching, I came up with:
MR=0.715
WR=0.5SR. (the spring compresses 0.5" for every 1" of wheel travel)

Therefore we, need a 525lb/in spring to handle a 1.5g bump in 4 inches.

Suspension travel has a dramatic impact on spring rate. In this example, if you lower the car another inch you would need 700lb/in springs – which is in fact what I did.

Suspension geometry also has a dramatic impact. If we use an eccentric bushing at the lower strut mount that could move the strut mount outward 0.5”, it would change our SR formula to WR=0.6SR. Now we could lower the car to 3” of travel and use 575lb/in springs.

Here I'm setup to measure the rear. I like to attach strings to all the suspension pivot points, then tie a weight to the end. Makes measuring a bit easier.



Thru the same analysis, I ended up with 500lb springs in the rear (and 700lb in the front). I think it has worked out pretty well. YMMV

Cash

_________________
09 Bullitt - DD
11 Fusion - Wife's DD
03 Mazda6 - Track car
00 PSD Excursion - Tow rig
67 Firebird - Don't you still own your 1st car?
61 F100 - Dad's truck
90 Moto Guzzi Calif III - Tourer
00 Cagiva - Adventure bike
00 DRZ400 - Woods bike

2.2/2½ Hz sounds reasonable, but why would you discard the sway bars entirely?
A - they are useful tuning tool
B - the weight penalty isn't that great and it is low down
C - with stiffer springs, the usual drawbacks of sway bars are minimized

Glad you are looking at suspension frequency and it really is the only way to get the real story.

I took my math for WR, MR, etc, from the Eibach suspension worksheet: http://eibach.com/america/en/motorsport ... -worksheet and got my weights from putting the car on the scales, then putting the unsprung components on a scale (I happen to have a bunch of torn apart Civics sitting around). I used 50% of the weight of the control arms, but didn't think about the axles. I'll have to weigh them. My springs are at 8 and 12 degrees, so my ACF is negligible.

I have a lot of ride height and suspension travel, so hitting the bump stops isn't an issue for me. When I was setting the ride height; my objective was to minimize dynamic toe and dynamic camber, so I set it so the tie rods are parallel with the ground and the rear upper control arms are at a slight down angle. Surprise surprise, that put it at right about stock ride height. At one point I modeled my suspension geometry in 2D (cardboard) and figured out my roll centers and got my roll axis parallel, but that was a lot of work and I thought that it was more important to have my control arms describing the correct arc than matching my roll center heights.

Cash, it looks like you're going at this from the perspective of static deflection, which I think is a slightly different way of getting to a workable result. I went at it from a natural frequency perspective, but it's easy enough to convert from one to the other once the measurements are done. After I get home (I'm still at the office) I'll plug your numbers into my worksheets and see how it compares to what I'm thinking about for the Duff car. Your car feels very good in corners, much better than mine, although I think you have more rubber under it than I do.

My question concerning static deflection is why stipulate 1.5g? I'd think that I want to corner until my inside wheel comes off the ground if I had the traction, so I'd stipulate 2+g, assuming that one inside wheel will get lighter before the other, so you'll actually start transferring weight longitudinally as the weight on one of the inside wheels approaches zero.

I'm ditching the sway bars because they are huge, almost an inch thick, and make a massive difference when connected. I think that Acura used them so they could put weak springs on the car for comfort, but still get it to corner. I'd like to get the car close to where I want it with springs and then do minor tuning with tire pressure. I'll keep the bars on the bus in case we want to use them, and they only take a few minutes to install. Maybe we'll turn some laps with and without them at the test day before the April race at CMP.

_________________
"Every ship is a minesweeper, once."

Couple of random thoughts:

(1) going with stiffer springs and no bar might make it so stiff in pitch as to unsettle the car and make it difficult to put power down over irregular surfaces. Are the cars with similar rates you mentioned also Civics with similar geometry, WR and MR? Are they sans sway bar also? What contribution to roll stiffness are the bars currently giving? Are you making up for the loss of roll stiffness with the stiffer spring rates, or is the car net softer than the prior spring/bar combo? If you are going to ditch the bars, you need to see how much impact that will have on roll stiffness in cornering because you could end up with a car that corners worse, AND performs poorly on straights as it jounces over track irregularities from the stiffer springs.
(2) 1.5 lateral G is pretty aggressive for 200TW tires, so I wouldn't be concerned about cornering deflection above that unless you have documented sustained lateral G forces above that (Not peak G), or moving to stickier tires.
(3) Keep the bars installed on the car and just remove one of the end links. Cuts down on install time if you choose to add it during a pit stop or test session. It takes me 7-10 minutes to do that on the rear of the S2000, and I am not in a hurry when I do that.

edit: In re cornering until wheel comes off ground. That could easily be a function of limited droop travel. I assume you are talking about lifting a rear wheel on corner entry, right? That puts a lot of cornering load on OF tire, which it may or may not be able to tolerate for sustained corners. Good for autocross, but may not be optimal for enduros.

_________________
Steve Carter
1972 Datsun 240Z-- resto pics at http://picasaweb.google.com/srcartermd
2007 GPW Honda S2000-- STR 86

Hmmm. We're running cut stock springs on crappy shocks. - AB

_________________
'14 Toyota Sequoia Platinum 4WD
Super Westerfield Bros - '93 Integra - LeChump Du Jour
STX 93 - Scion FR-S

Steven, the rates that I am considering are common on Civics and Integras with identical suspension geometry to the Duff car. Many of them have no roll bars.

The only cases that I can make for a roll bar are for a street car where you want to keep the ride smooth while combating body roll, or for a car that uses aerodynamic downforce to to push the nose down so you can get the nose low in the straights and not scrape in turns. They can be used for tuning, but I think that mine are too thick. Guesstimating the numbers because I'm at home now and doing some sway bar math, the smaller rear bar has a rate in excess of 430 lb/in, and the mechanical advantage of the mounting point is less than 2:1, so it would exert hundreds of lb/in on the suspension. The front bar is thicker.

The 1.5g and 2g numbers refer to vertical g's as they affect spring deflection and include the 1g that we get from gravity. The Duff car can barely exceed 1g laterally.

_________________
"Every ship is a minesweeper, once."

Using Cash's numbers to calculate his suspension frequency, I'm coming up with 2.2 in the front, which is funny because we each used a different methodology to get to the same number. Having ridden in his car, I'll be happy if the Duff car gets anywhere near that kind of handling with some stiffer springs.

_________________
"Every ship is a minesweeper, once."

Heh. Disconnecting the sway bars as it is now would put it back to the Stevie Wonder of days past. Me and you like that. Robert would think the car is broken. And Charles would likely go flying off the first corner not knowing to advance his inputs. :-)

Unless you're attaching rockets to those springs, no amount of tweaks will make Duff Class A fast.

Our front suspension geometry is different so there is a bit of apples/oranges. Mine is double wishbone - although it has a great camber curve it is quite limited in suspension travel. This is the reason I've assumed 1.5g bump. - knowing I'm at the low end of the spectrum. On a smooth track with a smooth driver 1.5g is fine, but if I start jumping gators, I'm way over 1.5g and on the bump stops. If I designed for 2g, my spring rates would be even higher or I would need to raise my ride height - like I said, its a compromise.

I even modified (raised) my upper shock mount to give me another inch of travel:






FWIW - I had to go to a much smaller RSB after increasing the spring rates.

Also, when I replaced the subframe bushings with solid aluminum, I made the flange a little thinner - this draws the frame up closer to the body, which raised the inboard end of the control arms in an effort to get them more level after lowering the ride height. They are still not quite level, but every bit helps.




Cash

_________________
09 Bullitt - DD
11 Fusion - Wife's DD
03 Mazda6 - Track car
00 PSD Excursion - Tow rig
67 Firebird - Don't you still own your 1st car?
61 F100 - Dad's truck
90 Moto Guzzi Calif III - Tourer
00 Cagiva - Adventure bike
00 DRZ400 - Woods bike

Actually Civics (before 2001) are double wishbone in front and have a trailing arm with LCA/UCA in the rear, so very similar to your suspension. I like the shock extension. I have something similar on my old Civic track car that I never drive any more, and I figured I'd move them to the Duff car, but they are machined aluminum and obviously aftermarket (https://www.jhpusa.com/store/pc/viewPrd.asp?idproduct=10197), so I didn't want them under the hood at a LeMons BS inspection. Something like you did wouldn't raise any eyebrows though.

I was thinking about subframe bushings for the Duff car. Civics and Integras don't have them; the subframe is bolted flat to the chassis, and only at 2 points. You just make sure that the big subframe bolts are extra tight to keep the whole thing from moving around. Sometimes they work loose and the whole front end shifts and clunks around under hard braking and cornering. I tighten them before events to minimize that, but before we realized what was causing it, the front of the car would jump around like the powertrain was trying to escape from the car when the bolts got a little loose. We went 2-3 races like that before we realized that it wasn't just a byproduct of the collision that sent the car our way in the first place. There is a company that makes machined aluminum bushings that positively locate the subframe on the chassis. I keep thinking about buying a set, but other priorities always intervene:

http://bloxracing.com/shop/index.php?route=product/product&path=63_110&product_id=245

_________________
"Every ship is a minesweeper, once."

FWIW. PTE Miatas generally run stock sways due to the way the classing works. In order to get the same roll resistance up front you end up putting a 1000# spring on the front of a Miata. It is generally considered just as fast as the more typical 700# with big says setup, but much more difficult to drive with considerably less forgiveness to mid-turn corrections.

1000 lb/in springs on the front of a Miata would give it a 2.86 Hz suspension frequency, stiffer than what I'm planning, but in the same 2-3 ball park. The Miata's Motion Ratio is a .69, vice a .81 for the Civic, so they need to have stiffer springs to get the same wheel rate.

The mid-turn correction factor shouldn't be an issue with a Civic. I took an instructor clinic with a Spec Miata racer and we both ended up driving my car for the exercises because his had no passenger seat. He was a bit alarmed the first time around VIR with my lavish, indiscriminate use of throttle pretty much everywhere, but that's FWD. If I get in trouble mid-turn, then I give it more throttle. He told me that Miatas aren't like that, which makes sense given that they put the drive wheels on the wrong end of the car.

_________________
"Every ship is a minesweeper, once."

That's what you gotta love about FWD - the solution to almost every problem is more throttle!

Cash

_________________
09 Bullitt - DD
11 Fusion - Wife's DD
03 Mazda6 - Track car
00 PSD Excursion - Tow rig
67 Firebird - Don't you still own your 1st car?
61 F100 - Dad's truck
90 Moto Guzzi Calif III - Tourer
00 Cagiva - Adventure bike
00 DRZ400 - Woods bike

I called yesterday about getting some new springs, and they told me that I should really get my shocks revalved if I go over 320lb/in. That adds about $500 to the cost of this, but given that it's basically the same cost as getting a new set or getting a set rebuilt, which I would have done in the next year anyway, I can live with the expense. Interestingly, Bilstein wanted to know car model and spring rate so they can do the math and match the valving to the suspension frequency, which should give me better performance and increased longevity out of the revalved shocks. I boxed up an old set of shocks and dropped them off at FedEx yesterday.

_________________
"Every ship is a minesweeper, once."