claykos

The formula for lateral load transfer from Milliken (a respected engineering text on vehicle dynamics), pg 683 (I am going to leave out the terms relating to gravity and the load transfer through the roll centers) is


delta_wf/A=(W/tf)*(H*Kf/(Kf+Kr)+roll center term)

delta_wf=load transfer across front axle
A=acceleration in G's
tf=front track
W=total weight
H=vertical distance from the roll axis to the center of gravity
Kf=front roll stiffness (lb*ft/deg)
Kr= rear roll stiffness

Same formula for the rear only with H*Kr...

Remember this is STEADY state cornering with no front to rear acceleration (which rarely happens on the track). The formula tells us that in steady corner there is no transfer front to rear, only from inside front to outside front and inside rear to outside rear.

It tell us that if we increase the proportion of the total roll stiffness at one end of the car, more load transfer will occur at that end of the car and the tires will be less evenly loaded, which due to tire nonlinearity will mean less total grip at that end. Conversely, the other end of the car will have the tires more evenly loaded and therefore more overall grip.

The way I think about this is as follows:

In steady cornering on a smooth surface with a rigid chassis, the car rolls some amount, which is the same across the front and the rear, and is proportional to the total roll stiffness (Kf+Kr in the above formula). The stiffer end of the car creates more resistance to roll for that given roll angle, and thus transfers more load.

Larry Herman

Yeah ^^^ that.

The math makes my head hurt, so I tend to think of things (and verify them) empirically. SO, based on that statement, if you stiffen the back of the car a little, more of the total load (that is the key item - total load) will be transferred across the diagonal from the front to the rear. We are moving a little grip from the back to the front, not just reducing grip in the back due to increased roll stiffness.

Now if we factor in things like improved geometry due to less lean angle, and a shift towards more equal contact of all tires, we actually can improve the total grip of the car somewhat by increasing the roll stiffness, with-in limits.

SundayDriver

Let's see if I can express my views without being able to draw it out.

Start with a car with 50/50 F/R weight and perfectly corner balanced. The total load weight transfer to the outside is a function of track (let's assume F&R track is the same), the height of the CG and the G load. Spring rates, roll resistance, etc have NO impact on this total weight transfer.

Assume the above geometry and cornering speed creates a transfer so that the balance is 60% right and 40% left (left hand corner). If the roll stiffness is the same F/R we still maintain 50/50 weight F/R and have not created any cross weight change (Wedge). If the car was neutral to start, it is still neutral.

Next step in the thought process is to split the car into two sections F/R and stiffen the rear roll resistance. This is an imaginary car so each end gets the forces as if they were connected but the don't interact with each other. Since the rear is stiffer, the front rolls more than the rear. But that does not change the weight transfer, only the amount of roll.

Now let's connect F/R back together and we see that the front is now acting through the chassis to try to roll the rear more than when they were disconnected. The twisting action is now trying to lift the LR and is resisted by the chassis stiffness. This coupling between F & R creates a wedge situation twisting the rear to put more force on the RR.
BUT - the total weight transfer cannot change, which is what creates the wedge. A LF/RR wedge will cause oversteer in a left hand corner. (Remember this is not about total weight, only cross weight at this point.) We are still at 60/40 side to side, but we have biased the diagonals. THAT is what I think explains WHY we get oversteer.

I hope that was reasonably clear and didn't make too many heads explode but that is the best I can do without a whiteboard.

Larry Herman

Very clear to me. Here's a simple test. Take your iPhone (or something similar) and pretend it is the car. With the phone longways, hold each end with your thumb and forefinger on each corner. Lift up your thumbs, tilt it up on one side slightly to simulate a car leaning. Now gently lift the fore-finger that you imagine is outside rear corner. See how it immediately pressures the inside front, simulating wedge across the diagonal. This is what IMHO transfers load from the inside rear to the inside front, reducing grip in the back and increasing grip at the front.

claykos

Yes, basically the total load transfer is just a function of geometry, like SundayDriver said. Cg height, weight, track. But how much of the load transfer happens at the front versus the rear is a function of how stiff the front is relative to the rear. So if we stiffen up the back, we transfer more load at the back and less load at the front. Therefore the front has more grip and the rear has less.

And as you said there are A LOT other factors, and remember this all assumes steady state cornering which almost never happens in a race car.

Add in shocks which effect the transients (speed of the weight transfer) and things start getting complicated quickly...

In reality, things sometimes work in reverse if you are far outside of the optimum operating window of the suspension/tires/etc. IE, you are way screwed up with bumpsteer, camber gain, etc and stiffening up the car gives you a lot more grip. For example a lot of older cars which had poor geometry, stiffening up the front gives you a lot more front grip because the suspension geometry and camber curves are so poor.

winders

Larry,

No.

Read this again:

Performance Handling by Don Alexander has this text in it:

We know that weight transfer while cornering reduces traction because of the way tire traction responds to vertical load. So now, if more weight is transferred at the front and less at the rear, the front tires will have slightly less traction and the rears slightly more traction. If we started with a neutral handling balance, the car will now understeer. The opposite applies if we stiffen the rear roll resistance only.

It does not matter if the increase in roll resistance comes from springs or antiroll bars. The effect is the same, although the manner in which weight transfers is different.

Again, increasing the vertical load on a tire does increase the traction of the tire but not in proportion to the amount of load. Not even close.

Scott

winders

Larry,

Roll couple is what determines how the load is distributed front to rear when cornering. "Roll Couple" is also called "Total Lateral Load Transfer Distribution" (TLLTD).

This article by Bryan Hise (JRZ USA) is interesting:

http://features.evolutionm.net/print...hp?cat=&id=109

Scott

SundayDriver

I don't agree that this is necessarily true as it depends on the tire and where you are on the load curve to start with.

Case in point, the last car I ran (in my avatar) would generate ~1.5 times vehicle weight with driver in downforce as speed. So the tires would go from 1X load to 2.5X load due to aero. Yet the cornering speeds, considering downforce, were not far off the expected with a constant grip from the tire. So in this case, there was very little fall off in grip as load increased. There is certainly a point where the load would hit the 'knee' of the curve and grip would fall off quickly, but it was not in the operating range even with all the aero load.

Some cars will be running near the knee of the curve without aero, and what you say will be true, but I feel it is dangerous to generalize. This is especially true because so many track/race cars run with very large tires to ensure that there is not a severe drop in grip.

BTW - since you are pretty new here, it would be nice to know something about your background and experience, what kind of cars you have worked with, etc. That would help in understanding what you are saying and put it into context and most of the posters in this thread have been here a long time and we have learned the backgrounds of each other over time.

On another note, it would be great is sometime we could actually get a bunch of us in the same place to really discuss this stuff instead of only doing it in the limited environment of the Internet.

winders

Mark,

The assumption here is that we are trying to corner near the limits of adhesion for the tires.

Aerodynamic downforce is a whole different ball of wax. You get the vertical load without the corresponding lateral load. So let's no consider aerodynamic downforce here.

If the load didn't affect grip as suggested, 2000 lbs and 4000 lbs cars with the same size tires, same CoG, same roll, same roll couple, etc. would negotiate a corner at the same maximum speed.

Regardless, even if you are not the limits, the traction gained by increasing the vertical load does not offset the increased lateral load. This is why stiffening the front suspension (spring or sway bar) reduces its grip.

Again, from Performance Handling by Don Alexander:

When the body of a vehicle rolls while cornering, the springs and antiroll bars resist the body roll. Stiffer springs and bars reduce body roll, but not the total amount of weight transfer. If we increase the roll resistance at one end of a vehicle only, more weight will be transferred at that end of the vehicle and it will lose traction compared to the other end of the vehicle.

Scott

claykos

I think the key is here, is not only will the stiffer end lose traction because it has increased load transfer, but the opposite end of the gain will gain traction because its load transfer is decreased! I think this is what Larry is trying to say about diagonal weight transfer, even though we do not actually transfer any weight diagonally when cornering. However if we compare the weight transfer before and after stiffening one end, then there has effectively been weight trasnferred diagonally (when comparing the corner weights before and after stiffening one end of the car).

winders

Good point. That is something that makes sense but is not always obvious to people.

Has the weight be transferred diagonally? I don't see how. How could changing the roll couple cause weight to transfer diagonally in a turn?

Scott

SundayDriver

Fair enough on aero but my point is that the grip is not drastically decreased because the normal load goes way up, even though I am using a car that is not typical of what is driven on track. Here are the things I question:

I contend that a 2000 and 4000 lb car in your example WILL corner almost the same, provided the tires are sized for the 4000 lb car and both somehow get the same tire temps.

I do not disagree with the results of changing roll resistance. What I am theorizing is that this is really driven by changes in diagonal load or wedge, far more than any other factor. Have you played with wedge in a car? I have and in a 1000 lb car, setting as little as 25 lbs of wedge has a pretty drastic effect on handling in left vs right hand corners.

You make a lot of sense and, again, I am very curios as to your experience and background.

SundayDriver

I think if we were face-to-face, I could show you how this MUST happen when you have a relative difference between F&R roll stiffness.

BTW - This is a great discussion and certainly making me think about what is happening with the car.

Larry Herman

That is exactly my point. I'm just trying to establish a mental model of how this happens.

claykos

Scott...I am not saying that weight is actually transferred diagonally, it definitely IS NOT!

What I am saying is lets consider case A, with a certain front and rear roll stiffness. Then in a 1G corner we transfer a certain amount of load across the front axle, and a certain amount across the rear.

Now lets change the rear roll stiffness (make it stiffer), call it case B. Take the same corner, now we transfer more weight at the rear than before.

If we compare the corner weights between Case A and Case B, the weight across the diagonals is different.

Does that make sense?