mikey_audiogeek

L Cubed, good work with the math.

How did you allow for the compliance of the sway bar bushings?

Cheers,
Mike

Dimi 944

I have a pretty balanced setup right now on my S2 coupe: OEM 944 S springs (125lbs/in), 24mm TB, 30mm/19mm sways on konis and lowered a bit, 225/255 tires. With the rear sway bar at the middle position the car feels quite neutral in corners.

DarrenD

Amazing job with the math L Cubed, I am loving this post!

I absolutely love the way my 84 944 handles. Here's my current setup:

300 lb front springs, Koni yellows(full stiff), Turbo S front anti-roll bar. Stock bushings. Steel control arms.

28mm solid rear Torsion bars, koni yellows(middle setting), Turbo S rear anti-roll bar. Welmeister bushings. Turbo Aluminum rear suspension.

-2* camber, very slight toe-out. No crown correction. Rear specs I can't remember. Car is lowered 2" front, 1.5 in the rear. Car is around 2450lbs.

235/40/17 front tires, 255/40/17 rear tires.

Very stable on turn in, slight tendency to oversteer when pushed but overall very balanced.

MAGK944

Just understand that the most important part of any analysis is missing from any calculations you do on paper....THE DRIVER. When and how they use the three pedals on the floor and the shifter and how they approach, drive through and leave a corner.

A good driver can make a bad setup better and likewise bad driver can unbalance a good setup just by their actions.

Driver training and experience on any track or course accounts for probably 50% of how well a car will perform. It's difficult to dynamically analyze that unless you have F1 style telemetry and feedback of a driver on a track.

L Cubed

Mike, I didn't account for bushing compliance or the slight variation in motion ratio as the wheels move through their travel range because I don't have the tools to do that measurement well. My assumption, and granted assumptions aren't that great, is that the compliance on the front roll bar bushings is similar to the rear roll bar bushings; meaning that they will effect total effective roll stiffness more than the balance or roll couple distribution.

I completely agree, it is almost impossible to predict the handling between different driving styles because of the way the driver reacts to what he/she feels and then adjusts their input. So finding the roll couple distribution that works is more a prediction of balance at a constant speed in a constant radius turn, like mid-corner or sometimes called "quasi-steady-state". In corner entry and corner exit I think the driver's preference and typical style of steering and pedal inputs dictate the perceived balance of the car.

Not that it means anything but I like to trail brake which typically tends to create oversteer into the corner and then have some understeer on the way out so that I can feel the limits of the tires as I try to exit with as much throttle input as possible. So I aim for a neutral "quasi-steady-state" setup and try to use the shocks and alignment to subtly adjust the transient entrance/exit behavior of the car. To be honest though, I have a lot of experience with formula autocross cars, a few years in autocrossing full body street cars, and no experience, yet, on a road course. So I am here to learn as much as I am here to share my knowledge.

L Cubed

I crunched the numbers from Dimi's and Darren's cars:

Dimi
Roll Stiffness's of:
Front = 1399 lb-ft/deg
Rear = 1199 lb-ft/deg (middle bar position)
Total = 2598 lb-ft/deg
Roll Couple Distribution = 53.8 % front

With 225/255 tires and 50/50 weight distribution, your "Grip" distribution is 46.9 % front. What is interesting here is that the 6.9 % difference biased towards the front, I would have guessed that your car would tend to understeer at limit. When we consider Arominus's car with a 50.2 % front RCD and 45.9 % front "Grip" that was reported as having understeer. Maybe there is something more to consider for the handling balance...

If you put the rear bar on the stiffest setting, that would generate a 48.4 % front roll couple which would be much closer to the 46.9 % front "grip" distribution. Does this setting also feel neutral Dimi or is it a oversteer, "tail-happy" setup?


Darren
Roll Stiffness's of:
Front = 1265 lb-ft/deg
Rear = 1306 lb-ft/deg
Total = 2571 lb-ft/deg
Roll Couple Distribution = 49.2 % front

With 235/255 tires and 50/50 weight distribution, your "Grip" distribution is 48.0 % front. Being around a 1 % difference and being described as neutral matches my hypothesis/prediction for how roll couple distribution and "Grip" distribution relate to handling balance. What size rim widths are using front and rear?

Dimi 944

Thanks for the analysis. I have not the pushed the car with the new tires yet. But so far it does feel pretty neutral. I will test it out in the next few days to see how it behaves at the limits.

Btw, I have a set of 27mm torsion bars if anyone is interested.

J1NX3D

interesting stuff! I'm about to transition. Are you able to calculate please?

I have an '86 n/a.

Current setup:
front- stock springs and shocks, 26.8mm sway bar (with braces), -2 deg camber, 225/50/16's
rear- stock torsion bars, 18mm sway bar, 245/45/16's

Eventual setup:
front- 200lb springs, -1.5" ride height, koni yellow sports shocks, 26.8mm sway bar (with braces), -2 deg camber, 225/50/16's
rear- 25.5mm 968 torsion bars, -1.5" ride height, koni yellow sports shocks,18mm sway bar, 245/45/16's

Please note I'm going from early offset to late offset, from '86 n'a suspension and brakes to full 968 engine, drivetrain, brakes and suspension if that affects anything other than weight.

L Cubed

Dimi, I would love to hear how it feels, especially if you get a chance to run at the different rear bar positions.

J1NX3D,
To be clear, this math is very much an educated guess. I do think were this does work well is as a comparison like what you are looking for. Unfortunately I cannot help with how lowering will affect the car, that is the next project I plan to tackle for understanding of the 944 platform.

Roll stiffness calculations:
Front (current / new) = 890 / 1001 lb-ft/deg
Rear (current / new) = 1003 / 1119 lb-ft/deg
Total (current / new) = 1893 / 2120 lb-ft/deg (+12 %)
Roll Couple Distribution (current / new) = 47.0 / 47.2 % front
"Grip" Distribution = 47.9 % front

So your new setup should be close to your old setup in handling balance and it looks like it should be pretty close to neutral!

DarrenD

Thanks L Cubed. I run 7.5" wide fronts and 9" wide rear rims.

When I decided to run the springs and T-bars I have I was worried the car would be harsh. It's not. And the faster I go, the better it gets.

Next up for my car is full sphericals. I'd like to eliminate any bushing deflection/binding. I previously built a car(CRX) with full sphericals, race valved Koni's and crazy spring rates and the car rode, and handled beautifully. Allowing the suspension to glide through it's travel did wonders for handling and ride quality. It wasn't nearly as harsh as you'd think.

There really is nothing better than a well setup suspension. Well, that and a roots supercharger...

Keep up the great work L Cubed.

L Cubed

Hey Darren, your setup is one I was planning on going towards later on, so it's great to here how it works. One thing I was thinking of though is going with Ground Control double adjustable Koni's, do you find the Koni Yellow's work well for this setup or are they at/near their "limit"?

Can't disagree with this!

SpeedyC2

L Cubed -

You are correct, my tires are 225 front, 255 rear. The car is lowered a fair amount. I run the rear sway bar in the tightest setting.

The spreadsheet I used assumes the car is in a 1-g turn, not stationary. It's not really a dynamic model, but it does assume loading and weight transfer. The short answer is my numbers and yours are not apples to apples.

FWI(may be)W, the spreadsheet I used reports, while in the 1-g turn:

Total Front Wheel Rate: 846 lb/in
Total Rear Wheel Rate: 309 lb/in

% Roll resistance front: 67%
% Roll resistance rear: 33%

Total Outer Front Tire Load: 1519 lb
Total Inner Front Tire Load: -20 lb
Total Outer Rear Tire Load: 1234 lb
Total Inner Rear Tire Load: 265 lb

Roll angle: 1.69 degrees

What I can tell you is I have seen some confirmation of the negative load on the inner wheel with the spring on the front strut sometimes "popping" out of the hat in a hard turn and causing some rubbing on the shaft on the strut.

L Cubed

Speedy, it is interesting to see how your numbers differ, would you mind sharing how you calculated them?
Based on the rear bar being on its stiffest setting, I calculate a similar 1.67 deg body roll (based on 18 in CG height and 3000 lb car weight) though with a 37.2 % front roll couple distribution (roll resistance on the front) and 2692 lb-ft/deg total roll stiffness.

This is a good lead-in to what I was planning to post, the equations I used to get these numbers since others seem interested in doing their own calculations.

Roll_Stiffness_from_Springs [lb-ft/deg] = ( Wheel_Rate [lbs/in] * ( Track_Width [in] )^2 ) *pi/180 * 1/2 * 1/12
The pi/180 converts radians to degrees, the 1/2 accounts for there being two sides and the 1/12 converts from inches to feet. The wheel rate is the spring rate at the wheel. To convert a spring rate at the strut to the wheel rate you multiply by the motion ratio squared. The motion ratio for the front I was using is 0.92; thus a 200 lb/in spring times 0.92^2 = 169 lb/in wheel rate. For the torsion bars, I used the numbers from Clark's Garage and Paragon tech site.

Using the classic mechanics calculation for Torsional Load, one can determine the torque required to twist an anti-roll bar 1 degree. From here you need to calculate the anti-roll bar motion ratio compared to body roll in degrees/degrees. For the OEM style front anti-roll bar, from my on and off-car measurements, I calculated the motion ratio to be 0.240 degrees body roll per degree of roll bar twist. For the rear it was based on the 19 mm 3-position with 0.134, 0.119, 0.103 (soft to stiff).

Roll_Stiffness_from_Roll_Bars [lb-ft/deg] = Roll_Bar_Torque [lb-ft] / ( Roll_Bar_Motion_Ratio [deg/deg] )^2
This roll bar torque is based on 1 degree of twist at the roll bar.

Adding the contribution of spring and roll bar roll stiffness for each axle allows the calculation of the roll couple distribution
Roll_Couple_Distribution [% front] = Roll_Stiffness_Front [lb-ft/deg] / ( Roll_Stiffness_Front [lb-ft/deg] + Roll_Stiffness_Rear [lb-ft/deg] )

This means that there can be some error in the actual roll couple distribution of the car under load if some of my measurements were off.

L Cubed

I realized after writing the previous post that I had made an error in my spreadsheet for the 30mm front roll bar, giving it the wrong motion ratio. This means when I calculated the numbers for Dimi's and Arominus's car

With the correct front roll bar motion ratio, the roll stiffness's are:
Front = 1170 lb-ft/deg
Rear = 1199 lb-ft/deg (middle bar position)
Total = 2369 lb-ft/deg
Roll Couple Distribution = 49.4 % front

Which means his roll couple is much closer to the "Grip" distribution of 46.9% than originally calculated which I believe would be closer to neutral as Dimi reports.

Corrected Roll Stiffness's are:
Front = 1278 lb-ft/deg
Rear = 1494 lb-ft/deg (stiffest bar position)
Total = 2772 lb-ft/deg
Roll Couple Distribution = 46.1 % front

This roll couple distribution is very close to the "Grip" distribution of 45.9 % front. Though the reported issues of some understeer and lift-oversteer might be related to other setup variables which have influence over the transient behaviors. That being said the 27 mm torsion bars could still be a fair change, creating a 47.8 % front roll couple distribution with the rear bar on the middle position.

mikey_audiogeek

I may have missed it, but have you incorporated roll couple due to roll centre height?
Given the difference between front and rear rc height, this has a significant effect.
Cheers,
Mike