What does a "race" suspension really do? 3

[1Sick6]

Hi, pretty sure this is my first post, but i've been reading on here for a few years - great forum!

Ok, I've read tons about the details of suspension design (ie roll rates, roll stiffness, roll couple %...) but as I took a step back and was trying to explain the big picture to someone, I realized I don't really know what's going on haha.

Basically, my question is what is the goal of a "race" suspension?

From what I've absorbed (through both good and lots of bad info) the two goals are:
- to keep as much of the tire in contact with the ground for the greatest amount of time possible.
- to limit "weight transfer" in order to get the most out of the tires.

I realize there are tons of other factors that go into the design, but I'm just trying to figure out the basic goals that are directly related to how fast a car can go around a corner. Am I missing any?

The first goal makes sense so long as you stick to the convention that more tire on the ground = more grip (something that my physics teachers try to challenge but I'll keep ignoring them for now)

The second goal also makes sense, but its what i've got the most questions about. If load transfer is only related to mass, CG height, and track width - how do things like ARB's and stiffer springs result in more grip? How does the "all evil" body roll hurt performance? The same load is being transferred through the springs (and ARB's), they're just not compressing right?

My only thought on it is that as the body rolls, the CG is shifted slightly to the outside of the turn giving us *some* weight transfer (we actually get to use "weight" for once). The part i struggle with is: how much does this really affect things?

Also, I hear people talking about "balancing" the car with ARB's. Saying that "a stiffer bar will reduce load transfer at that end of the car." Then they get into roll couples and then appears all kinds of facts about car behavior due to one end of the car resisting roll more than the other. Where does that come from?

Sorry this ended up so long,but if nothing else I already better understand whats going on by typing/thinking this out. Thanks for the help!

-matt

 

[TrackRat]

The objective of a race suspension is to obtain optimum performance from the tires, which generally means optimum traction from the tires.

Weight transfer is in fact fixed based on the mass, CG and trackwidth and acceleration. How you deal with the weight transfer determines how effective the tires can function.

Tires generally produce the greatest traction when they are flat on the road surface, aka zero camber. When the suspension moves the camber of the tire changes. When the vehicle body rolls the camber also changes. Depending on the design of the suspension, you can minimize the camber change or make the camber more negative to compensate for body roll. Suspension movement and body roll are controlled by springs, swaybars, dampers, etc.

By optimizing the spring rates you can effect how well a tire grips and the handling characteristics of a vehicle. A swaybar is just a torsion bar which is a spring. A swaybar or stiffer spring can be used to reduce body roll. By using a stiffer spring or swaybar you apply more load to that particular tire or axle. This changes the tires traction and the vehicles handling.

A good book on vehicle dynamics will show you how to apply the basics. Herb Adams' book "Chassis Engineering" is one of the easiest to understand in my experience. He starts with the basics and walks you thru the process so that you understand how it's all related. There are many other books that can help including Fred Puhn's "How to make your car handle" and Carroll Smith's numerous engineering books.

 

[1Sick6]

Thanks for the reply! I'll have to check those books out. I've read Valkenburgh's "Race Car Eng and Mechanics" and then parts of Milken's RCVD - quite a jump haha.

Everything you said makes perfect sense except: "By using a stiffer spring or swaybar you apply more load to that particular tire or axle."

I thought the load came from the mass, cg, track and acceleration? I'm not seeing how a stiffer spring could increase the load a particular tire.

I have a decent grasp on how a suspension works as far as the geometry is concerned, but I feel that I'm missing the big picture. - how does the suspension control load transfer?

 

[TrackRat]

The sprung vehicle mass is transferred to the tire as load via the spring. A stiffer spring applies more load to a specific tire. By changing how much load a particular tire has during lateral acceleration, you determine which tire has the most traction. That is why changing a swaybar or spring changes the handling, by changing the load on a specific tire during cornering. By changing the loads on the tires you can balance the work that the tires do during cornering to obtain as much traction as possible.

 

[NormPeterson]

A tire's grip is not linear with respect to its vertical load. That's the root basis behind tuning the understeer/oversteer balance via shifting the front to rear lateral load transfer distribution.

Best lateral grip probably comes when the operating camber is slightly negative rather than precisely zero.

BTW, it's better to think in terms of lateral load transfer rather then "weight transfer". The only weight transfer that physically happens in a vehicle comes from fluids sloshing around in their containers and insufficiently strapped-in people/cargo doing likewise.

It might be a good idea to stop listening to anyone who says that "a stiffer bar will reduce load transfer at that end of the car.". At least independently verify every single thing said.

There are quite a few things that affect cornering/handling beyond suspension stiffnesses. The suspension geometry will cause cambers, toes, and axle steer to shift a bit as the suspension moves. Other things like bushing compliances and suspension bracket flexibilities will "give" a little under lateral (and longitudinal) loading and introduce what you might consider to be "errors" with respect to what the theoretical geometry would produce. These and a few other things fall under the topic of "understeer budget", which is sort of a sum total of all of the effects.


Norm

 

[YvesLLewelyn]

Speaking as a distinct non-expert in suspension matters - it seems to me that in recent times most of the race suspension setups are aimed at keeping constant distances etc. between the parts of the cars body/chassis and the road surface to maximise aerodynamic downforce. I think the drivers etc. put up with overstiff springs/odd handling etc. as the increase in grip due to downforce seems to be the controlling factor.

 

[1Sick6]

"The sprung vehicle mass is transferred to the tire as load via the spring. A stiffer spring applies more load to a specific tire."

Again, this makes no sense.

Assume there is a 100lb/in spring on the wheel on the out side of the corner. (only considering the front here)

Also assume the load transfer = (accel)*k where constant "k" is from mass, cg height and track width. So for a 1g turn the load transfer is k lbs. For simple numbers take k to = 500 lbs

So there is a constant load on the spring of 500lbs - this will compress the 100lb/in spring 5in.

So what is there is a 250lb spring? It compresses 2in. But the force is the same right?

How is the spring rate actually supposed to change the wheel load? I could see it causing the load to maybe act more quickly on the tire, since as the 500lb is applied to the top of the spring, the 100lb spring will not exert the full 500lbs until the spring is fully compressed to 5in but the 250lb spring will exert the full force in 2in of travel.

A tire's grip is not linear with respect to its vertical load. That's the root basis behind tuning the understeer/oversteer balance via shifting the front to rear lateral load transfer distribution.
I understand that if you can control the load transfer you an change the balance - i'm just missing how the load transfer can actually be controlled.

It might be a good idea to stop listening to anyone who says that "a stiffer bar will reduce load transfer at that end of the car.". At least independently verify every single thing said. So how does a stiffer bar effect load transfer? - and how does it do it? It seems to me that ARB's are just a way of raising the outside wheel's spring rate, but only in corners.

Thanks again for helping me sort through these questions. I've heard so many people say that doing X will change you load transfer like "this" - but I've never understood how the load transfer is really controlled.

 

[TrackRat]

The point you are forgetting is the chassis is hopefully rigid, so both the front and back of the chassis roll the same amount. When the chassis rolls the springs apply load to the tires. A stiffer spring applies more force to a given tire so that tire shares a greater part of the total weight transfer.

A stiffer swaybar applies more load to the outside tire as well as trying to lift the inside tire. You could apply the same load to the outside tire with just a stiffer spring but then the ride would be harsher all the time. If you only need to limit roll in corners then a swaybar is a convenient means to reduce roll and retain softer springs for better ride. A swaybar is a double edged sword, but effective within a reasonable range for many applications. Good suspension tuning is a lot of science and art combined.

 

[NormPeterson]

"The sprung vehicle mass is transferred to the tire as load via the spring. A stiffer spring applies more load to a specific tire."

Again, this makes no sense.

Depends on the kind of load.

Displacement loading due to roadway or track unevenness will affect the wheel loads for more stiffly suspended wheels more than for less stiffly suspended ones. From a purely theoretical structural standpoint, a four wheeled vehicle is statically indeterminate, meaning that except by wild cosmic coincidence or careful problem definition you need to know a few things about the stiffnesses in order to do the analysis. Kind of a long way of saying that you can't get away from considering stiffnesses.

Transient corner loading (i.e. while the suspension is moving from one position to another position) is more complex, as you then get to consider velocity-defined damper loads in addition to the position-defined changes in spring and sta-bar loads.


Norm

 

[1Sick6]

The point you are forgetting is the chassis is hopefully rigid, so both the front and back of the chassis roll the same amount. When the chassis rolls the springs apply load to the tires. A stiffer spring applies more force to a given tire so that tire shares a greater part of the total weight transfer.

This makes sense - I can see how a stiffer spring on one end of a car relative to the other end (ie front stiffer that the rear) would affect which wheel gets more of the load transfer. But i'm still not sure how the stiffness effects the overall amount of load transfer.

Assuming you have a nice 50/50 weight dist, the load transfer to the out side front and rear tires should be about the same. Balance aside, you need some way keep the inside tires loaded to increase overall grip - but will stiffer springs do this?

Does an ARB add load to the outside wheel by transferring it from the inside wheel? Wouldn't this be the opposite of what we want? I mean, it adds to the spring rate of the outside wheel, so body roll is reduced, but then the rest of the tires end up with less.

Kind of a long way of saying that you can't get away from considering stiffnesses.

Ok, I can accept that, but how do you consider the stiffness with respect to load transfer? All the statics i've done were simplified assuming no bending or motion so stiffness is a new factor to me in that regard. I'm missing what the stiffness characteristics add to the problem to fix the indeterminance. <- if thats even a word haha

 

[NormPeterson]

Does an ARB add load to the outside wheel by transferring it from the inside wheel? Wouldn't this be the opposite of what we want? I mean, it adds to the spring rate of the outside wheel, so body roll is reduced, but then the rest of the tires end up with less.

When you add a sta-bar (or replace an existing one with a stiffer one), the total lateral load transfer due to the roll remains the same. But you shift more of the LLT to the end that has been made stiffer.

Ok, I can accept that, but how do you consider the stiffness with respect to load transfer? All the statics i've done were simplified assuming no bending or motion so stiffness is a new factor to me in that regard. I'm missing what the stiffness characteristics add to the problem to fix the indeterminance.

You have to consider both front and rear roll stiffnesses for pure roll - and since they act in parallel the load transfer from the roll apportions itself according to their relative stiffnesses. That's part of it. Then you add the load transfers that don't contribute to roll (lateral forces through the "roll centers" and LLT's from the unsprung masses) to get (approximately) the total effects at each corner. Puhn's book discusses these three effects.

BTW, I think the word you're looking for is 'indeterminacy'.


Norm

 

[1Sick6]

haha - thanks.

I'm going to have to get a copy of that book... Until then - in general - can it be said that the overall load transfer from roll can be reduced with a higher spring rate at all four corners and everything else remaining constant?

 

[NormPeterson]

No.

 

[patprimmer]

The centre of gravity moves sideways with roll. The degree is generally insignificant in race cars as they do not roll much.

I am not a suspension guy, but I think that if the CG is above the roll centre the CG moves out and if below it moves in.

Regards
Pat
See FAQ731-376 for tips on use of eng-tips by professional engineers &

http://eng-tips.com/market.cfm

for site rules

 

[GregLocock]

CG lateral motion for any competent modern car is a tiny effect, maybe 10mm at 1g+tire deflection, say 50 mm. The wiki article on load/weight transfer was mostly written by a pedant with a theory.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[BrianPetersen]

The *total* left-to-right weight transfer (add up front and rear) must be the same (almost) no matter the roll stiffness. Changing spring rates or antiroll bar rates will not change this. Changing roll stiffness balance front to rear will affect which end is taking up more of the weight transfer and that will affect the understeer / oversteer balance by changing the way individual tires are loaded.

If you put a stiffer rear antiroll bar and do nothing else, it will tilt the balance towards oversteer. If you then put a proportionally stiffer front antiroll bar, it tilts the cornering balance right back to where you started, just with less body roll and less compliance.

But, as always, there is more to it than this. You need compliance in order to try to keep tires in contact with the ground and avoid upsetting the chassis on bumps. This inherently means there will be some body roll and reduces "feel" transmitted to the driver. Race cars that operate on near-glass-smooth tracks can get away with high spring and damping rates to limit body roll and maximize "feel" without losing grip much, but if you try to drive something like that over a typical normal potholed and broken road, it will bounce and chatter from one bump to the next and be almost uncontrollable.

I've seen on-board footage that showed suspension movement on an open-wheel race car on a track that has some undulations to it, and it was clear (to me) that they were using a very high antiroll-bar rate and relatively soft spring rates. There was next to no body roll in corners. This type of setup has its own set of issues because it allows compliance but essentially forces the left and right wheels to go up and down together, and you could see this on the video. A one-wheel bump with that type of setup tries to unload the opposite wheel and forces the car body to lean to follow in lean. If ride quality is a consideration, this causes an uncomfortable phenomenon called "head toss" on one-wheel bumps, but on a race car, that's not much of a consideration.

It is a fine balance between having the compliance needed to maintain grip on bumpy surfaces, but the stiffness needed to maintain "feel".

By the way, there's another, more common example of a suspension having a soft rate on two-wheel bumps but very high roll stiffness - the airbag suspension on transport trucks. That is an application where you cannot allow much body lean, because the center of gravity motion would be excessive and allow it to tip over more easily, so the linkage allows very little difference in motion between the left and right sides. Despite the very soft spring rates provided by the airbags, the ride is still not comfortable due to the head-toss phenomenon on one-wheel bumps.

 

[patprimmer]

Greg

So it is a hell of a lot more to do with compliance rather than roll? I was thinking only roll and therefore almost nil.

Regards
Pat
See FAQ731-376 for tips on use of eng-tips by professional engineers &

http://eng-tips.com/market.cfm

for site rules

 

[GregLocock]

Yup. You can generally keep the semitrack gain to less than 10%, and typically the wheel lifts by 50mm at 1g cornering.

However, watch out for camber gain, that could be quite a big effect.

Of course all this is ignoring the movement of the centre of pressure under the tire, , which I'd guess would migrate towards the outside of the turn, relative to the contact patch, but the contact patch migrates towards the centre of the turn.

Hmm.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[BrianPetersen]

It isn't compliance "rather than" roll. It is compliance AND roll control AND camber control AND controlling the way the bodyshell moves after hitting the bump AND maximizing tire contact with the ground AND balancing understeer versus oversteer AND controlling toe with regards to suspension movement and braking or bump or impact loads applied to the hub. Very very complicated.

 

[TrackRat]

^^^^^^ Damn I thought it was about going around turns fast... *LOL*

Is all this other stuff necessary?