F355 suspension analysis

As well it should.

My numbers used -0.8 as the front camber, so you are right on the mark here. Caster controls the fell of the left->right or right->left transistions. Its not overly sensitive until it get more than a full degree out of balance. You are running slightly less toe than I. Toe deadens the steering in the center and calms the car.

At the rear, I run -1.8 camber and -0.25 toe. Your rear camber and toe will cause the rear end ot be lively. If you like it this way, stick with it. If you want to calm the car down a little, add toe and subtract camber. My camber was finally set with a pyrometer on a road race track after some really hot laps. Getting the temperature profile nice and even across the tread while finding the correct air pressures.

If you are fighting the car on the street, or to rephrase the car won't let your mind wander/relax on the stereet--This implies you might want to take camber back to -0.5 front and -1.8 rear while dialing in some more toe to -0.25 both ends. This will deaden the cars on-center response without getting rid of the quickness, sensitivity, and balance with power on.

 

Yes, the pictures are accurate to a couple of hundredths of an inch over 70"
The stock ride height is confirmed in the measurements taken off the picture and with the factory shop manual.

When the roll center is at ground level, the forces in the suspension arms can change direction. The following graph shows the height of the roll center with respect to dive and roll for the front suspension: Notice how there are tow places where the suspension is 'confused'. One of the denominator terms went to zero. In actuality, one or more a-arms will changed from {tesnion:compression} into {compression:tension}. As this happens, the load on the bushings alters direction and the arm moves (a couple of hundredths of an inch) to take the new force vector.

 

If you have/want/need your roll center at ground level, you are far better off to put the roll center a little higher or a little lower. But because when the front suspension is doing work, it is in a compressed position. Therefore, you either put your roll center just below ground level (like -0.5" with F355) or just above 2" (C5 vette) but rarely anything in between (street applications).

This then interacts with the location of your rear roll center,.....

Other than leverage effects why do you want the roll center at any one point or another?

Background;
a) the less spring you can run the more tire traction you have
b) the less camber change the more traction you have
c) the lower the CoG the more traction you have
The only reasons to use stiffer springs are:
a) to avoid loosing more traction in camber than you loose in response to weight transfer
b) to avoid hitting the road surface

With the front roll center lower than the rear roll center, the front takes more of the forces in turn, and this naturally generates mild understeer, but more importantly, it positions the operating point somewhere forward on the doughnut of traction. It turns out that in braking it is useful to have the front suspension loose a tiny amount of traction as it nears lock up deceleration rates, promoting stability under brakes. Positioning the roll center below the road surface positions operating point such that this effect is within alignment adjustment range. One might even use <say> ride height as the adjustment mechanism.

The downside of using this kind of adjustment is that you pick up a twitch as steering is added under brakes. Careful selection of springs sets the traction circle as BIG as possible without transversing the twitch. This limits how soft the front end can be in the F355.

In addition, with the roll center so positionied, the car will naturally understeer at the front under acceleration while turning <the forward bite regison>. So as the rear drifts into power on oversteer, the front end slowly sheds traction enhancing controlability while driving sideways under power.

At the rear the roll center is higher (4.2" for F355) (~5" vette). The high roll center shortens the lever arm and transfers less weight through the spring and more weight through the suspension arms from to the outside tire, enhancing traction. Since the rear end is not so inclined to roll from the short lever arm, the springs can be less stiff and still control body roll, also enhancing rear traction.

At the rear the raised roll center also rotates the traction profile such that maximum traction is available across the whole power on corner exit phase of cornering. So, not only is traction maximized under power, but the fron end washes out at just the right time to prevent acceleration in yaw (spin) WHEN the car is setup correctly.

The downside of using this kind of mechanism is that you may run out of traction in pure linear acceleration if the rear tires overcamber in squat. This induces fishtailing. In general you like to set up the car with access to this operation region. TO have access, transmission gears are selected to put the available traction just above the available traction to give the driver feel under throttle. However, if the car is given a serious power adder, the driver may find himself in uncharted waters (similar to the lowering without alignment situation above)

With the front ride height set to avoid a roll center at zero, the rear ride height can be used to transfer weight from the rear to the front under cornering loads. This allows one to ajust the oversteer / understeer relationship without changing any parts on the car.

The reason modders of other cars go with spring and shock packages is to get the adjustability which arrived with the Ferrari. The stiffer springs and shocks fool the casual driver into thinking he has made significant forward progress in his handling, whether he has or not is often never known. The issue is completely confused when he also changes tires, wheels and brakes at the same time. Most of the time, any reasonable set of parts would make a similar step forward. This is more indicative of the factory setup than the tuner setup.

 

In the following picture, there are 3 points of intrest that distinguish the early F348 from the SS and later F348s and the subsequent F355.

1) the chassis pivot point of the upper a-arm was lowered
2) this decreased the distance between the chassis center line and the instant center
3) this caused the roll center to raise

The raised roll center moves the traction profile to a better position with respect tot eh suspension alignment.

In effect, the early F348 suffered from not enough rake in the roll center at the front with respect to the roll center at the rear. Thereby, not transfering enough weight to the outside front tire to wash out traction as power was applied in cornering. Making for tricky handling.

 

I should be clear here, I invented this way of looking at the suspension effects versus traction effects. I have never seen it in any of the multitude of suspension/handling books I have studied.

 

As far as ignoring deflection in the suspension bushings, control arms, hubs, and tires; I am guilty. At the level of this presentation, this level of detail would not enhance the readers understanding.

As far as ignoring weight transfer, I am also guilty. However, I don't know the 3D location of the Center of Gravity, and have no way of computing the weight transfer because of this missing data point. However, I do have a way of measuring the dispalcement that takes place at the wheels relative to the body, and the traction circles used are representative of the F355 in stock condition running on max performanc street tires.

I know you can move the kinematic points around to make the car roll out of turns and rise under braking. HOWERVE, I know of no car on the road today using these kinds of suspension geometries. If you find one, it is a simple job of flipping the traction circle in the correct direction or distorting the tob/bottom and left/right dimensions.

I agree with this. The method allows the causal observer to take a car that is already WELL setup and understand what small changes might do. It is not useful in setting a car up from scratch, nor does it supplant real measurements on real cars or more detailed simulations.

It is, however, good enough to explain why one should not lower one of these cars without some knowledge of what they are getting into. And that is the reason I put it together.

And it is good enough to explain what the driver feels in the transition from pure braking as one adds steering input, and why these cars put power donw so well exiting turns.

 

I hopw my response did not indicate any negativity with respect to your acknowledgements of the inadequacies of the technique.

What I did was to take the scans from the owners manual, and then in a CAD program scale and position those scans so that they represent both sides of the actual suspension. I measured a couple of points (on my F355) just be verify the scaling and positioning. From this drawing, I placed measurement points on all the important pivots. Then I took those measurement points to the 3D suspension analysis program.

The lines denoting the roll center and instant centers were only drawn for the benefit of the reader and were not actually used in the 3D analyzer, however, the 3D analyser DID come up with points that were within 'drawing' error of the drawn points (static position). And indeed, these points do move around as the car moves around on its suspension.

I don't actually remember what 3D program it was, as this data gathering and analysis was done over a year ago. I do remember that the software was a free download for 2 or 4 weeks of use. I removed the software when it was no longer useable. I have written a 2D version in eXcel, and that program agrees very well with the 3D program. One item I would like to point out to the curious, is that the program I used seems to be using 32-bit floating point numbers (instead of 64-bit or 80-bit floating point numbers) at least in a few cases. This can be seen in the traction profiles as jagged edges on the surface profiles.

I'll be waiting.

 

A) have you scrapped the nose with road surface anywhere in the last 6 months? If not, where do you live? drive? I can't even get out of my own driveway without touching the nose at stock height.

B) I should forwarn you that 'I' consider lowering a car to make it look agressive is 'RICE'. Lowering it for a purpose is acceptable when you know what you are doing.

C) Lowering a car such as an F355 (which starts out 1.5" lower than a Vette) does not gain as much as lowering a Vette since the CoG is already low and the track is wide.

D) a spider is not generally considered a track-like car, so the question of why you want to move the car in this direction returns to the concept of RICE.

E) adding spacers adds stress to the axle bearings, going for 1" of spacers on each side is very agressive and might require some recountoring of the inner fender area to avoid rubbing.

But on with the questions:

If you are going to lower the car 1" you should be looking at springs that are about 40% stiffer than stock, and you SHOULD change the roll bars to match this 40% stiffness gain*. If you lower the car 2" (you now only have 2.2" of ground clearance) you are going to need closer to 100% gain in spring rate to keep from bottoming. Shocks need to follow the spring rates. These spring rate suggestions are stiffer than needed by the traction profile analysis and are used more to keep the car from <ahem> interaction with the road surface than in optimizing the traction profile.

So lowered, the car will transfer weight significantly faster (lower CoG, stiffer springs and shocks) so the driver has to stay ahead of the car (no resting on long trips). Most drivers cannot do this with a stock suspension over 120 MPH unless they get regular track practice (like 6 times per year). The car will slide easier, and won't have quite as much traction due to the stiffer springs. Interior comfort will decrease and shakes and rattles will increase. Finally, the car will become very sensitive to suspension alignments and alignment procedures.

*Side note: why does every high performance car NEED anti roll bars?

Explanation: consider the displacement of the front suspension under two sets of acceleration a) heavy braking, b) pure cornering; both performed at 1Gs of acceleration.

The longitudinal dive under braking has to do with the weight transfer and the effective spring rate at the front wheels. This is governed by the CoG and the wheelbase.

The latteral roll under heavy pure cornering has to do with weight transfer and the effective spring rate of the side of the car. This is goverened by the CoG and the track.

Most cars have a wheelbase that is around 150% (+/-) of the track.

So the dive under brakes will only be 66% (1/150%) of the roll under pure cornering. What we want is to be able to transition from pure braking to pure cornering with the outside tire not raising nor falling during this transition. An antiroll bar set to the correct value adds 'effective' spring rate in roll that is not present in dive to accomodate this requirement.

So the antiroll bar is making up for the wheelbase being longer than the track such that the suspension movement at each corner is normalized in both acceleration directions.

 

RICE is what happens to innocent automobiles (mainly Japaneese) when enthusiastic (younger) owners spend all their cash and energy on wings, body work, paint, wheels,.... trying to trick out a car and make it LOOK fast (without actually making it fast).