Longitudinal weight transfer question

[adambrouillard]

Ok, I have a setup problem on my racecar. It's called a spec racer and it has equal size tires all around, 63% of weight (1670 lbs total) in the rear and about 100 horsepower and no significant aero.

I have noticed most drivers run more roll resistance in rear than in the front because at full throttle in all but the slowest corners you can not get power oversteer and they are feeling understeer. This however creates bad corner entry oversteer if you do any kind of real trail braking or anytime you have to lift mid corner. Most seem to brake in a straight line and then turn in under power or are constantly fighting oversteer at corner entry ( or not at the limit). This setup feels very unstable to me and before I give up and sell the car I am wondering if this is really the fastest way to drive a car like this.

My question is basically this. When the load transfers to the rear on corner exit, or in their case the entire corner, (data acquisition shows about .1 g acceleration in most medium corners.) does the load transfer actually add load sensitivity at the rear or take it away.

To put it another way, let's just say the rear of the car weighs about 1000 lbs and under steady state corner can achieve 1 g and has a combined lateral force of 1000 lbs. Now we transfer 100 lbs to the rear. The tires are now seeing 1100 lbs, but are they still having to push only the 1000 lbs or does the load transfer actually make them have to push 1100 lbs. In the first case the load sensitivity would increase and they might could achieve 1.1 g or some such whereas in the latter case they maybe could only achieve .9 g. I'm obviously fudging the numbers, but you should get the point.

I have been having a really hard time figuring this out. Normally this is not a problem because you typically have enough power to get the rear tires to the limit on corner exit on most cars. My instinct is to put a lot of roll resistance in the front to try and properly balance the car, I have tried this and it feels much more stable, but I am wondering if I am compromising total cornering power.

[Casper]

What works for you is the answer. Others may like that style of setup you don't feel comfortable so need to modify things to suit you. When you start doing things plot the time trends over a race distance not short one or two lap screamers. Driving is all about confidence and above all feeling good about the car. I bet at the moment you don't even want to drive it anywhere near a competitive speed.

[phoenix]

Not disagreeing with Casper at all, but I would add the following.

Lets deal with the weight distribution first: 63% rear and 47% front.

If you have equal roll stiffness front and rear, then for every 100 lbs transfered, 63 lbs will go on the rear and 47 lbs on the front.

It is clear, therefore, that as the tyres are the same on all four corners, the rears will be over-loaded before the fronts so the car will tend to oversteer anyway, without the input of any torque into the rear tyres from trying to accelerate while cornering.

Combine the two factors above and in your example of 1000 lbs being transfered, 630 lbs are added to the load of the outside rear and 470 lbs to the outside front tyre. The rear outside tyre is now carrying about 1156 lbs and the outside front 862 lbs of vertical load. Setup like this the car will tend to oversteer.

It should also be clear that just as the rears will be overloaded first, the capacity of the front tyres is not being used fully, and to do so would mean transfering more weight from the rear outside tyre to the front outside tyre.

With your car running on the same tyres all round, maximum cornering G will be achieved when the front and rear tyres are loaded equally and therefore slipping/gripping the equally. (This is slightly complicated by the fact that the front wheels are turning, increasing the slip angle, to make the car go round the corner, but let's ignore that for the moment)

To get equal weight on the front and rear outside turns, in your example, would mean sharing the 2018 lbs load (50% of static load + cornering load) equally front and rear - i.e. 1009 lbs front and 1009 lbs rear.

To achieve this, by my calculations, the front roll stiffness will need to be 57.3% of the total roll stiffness and the rear roll stiffness the other 43.7%.

With the car setup like this, and if the tyres were at the limit of adhesion, adding 100 lbs more load to the outside rear tyre from accelerating torque would cause the rear to lose some lateral grip and oversteer would set in.

I am not saying the above would be quickest in the real world, but I would start of testing with that setup (rather than the roll stiffness distribution you have) and then adjust roll stiffness distribution a few percent either way to see what is quickest and/or suits your driving style better.

We started with a 50/50 roll stiffness distribution on a car with a weight distribution of 57% rear and 43% front. We ended up after testing with 65% front roll stiffness and 35% rear for neutral handling. However, we have different front and rear track, which if your track widths are not equal you should bear in mind.....

[adambrouillard]

What you are saying was my inital thought, but I am actually asking how the longitudinal weight transfer effects the lateral capability.

To clarify, let's say we are transfering 1000 lbs total laterally. But we are also transfering 100 lbs from front to rear because of .1 g acceleration. Will this 100 lbs add cornering capability at the rear or take it away? My guess is it that it adds cornering capacity in the same way a wing would. I think it adds download without adding extra weight, but I have always wondering if this is true.

You are also using the rear tires to generate the .1 g of acceleration though so the net effect could be zero. However you are losing weight on the front and this would be a net understeer effect. This is of course assuming you do not have enough power to overcome the rear tires as is the case in my car.

[adambrouillard]

Ok, I think I figured it out. I don't know why I have been considering longitudinal and lateral separately. Just as with lat weight transfer having a net lower g capacity, it would effectively work the same way on longitudinal.

The extra weight would add capacity, but not at the same rate as the weight added so the net g capacity would be lower.

In other words, all these guys running more roll resistance in the rear are wrong, unless I am missing something.

The only way they could be accelerating through the corner is if the rear tire had some reserve capacity and they were not at the limit on the rear, and if the roll couple is biased toward the rear, then the front is certainly not at the limit and any perceived understeer must have been induced. Now my problem is that in order to get more than about 55% roll couple in the front I have to disconnect my rear bar which will hurt my overall roll resistance no a car that is already pretty soft.

[GordonG]

the longitudinal weight transfer caused by acceleration will add vertical load to the rear tyres, but will not require the rears to do any more lateral work. However, that is more or less balanced out by the additional grip requirement (assuming rear wheel drive) - in other words you are asking the rear tyres to accelerate the car longitudinally as well as laterally.

Whether the acceleration gives a net plus or minus to the rear grip depends on things like wheelbase, track, tyre grip and (probably most important) CoG height.

If the rear is too soft versus the front, you will only lift a front wheel in the en, at which point you cannot get any more load transfer at the front and balance tuning with relative roll stiffness stops working. however, 3 wheeling may be a perfectly good setup on some cars.

G

[phoenix]

Originally Posted by adambrouillard

Ok, I think I figured it out. I don't know why I have been considering longitudinal and lateral separately. Just as with lat weight transfer having a net lower g capacity, it would effectively work the same way on longitudinal.

The extra weight would add capacity, but not at the same rate as the weight added so the net g capacity would be lower.

If the car is using all the tyre grip available to corner, adding extra load to the rear tyres by acceleration will cause the loss of lateral grip - not give more grip.

[tristancliffe]

If the car is using all the available tyre grip to corner, and is then asked to accelerate as well it should lose traction. But the act of opening the throttle and transferring load backwards will increase the normal load on that tyre, increasing it's available grip, the same amount of which is taken up laterally. Thus the tyre now has some potential for accelerating the car a bit more, which gives a bit more capacity. Hence cars can accelerate out of corners whilst cornering at the 'max'.

Obviously this won't work if the car really is at the very limit of traction, but since even the F1 drivers aren't this good I wouldn't worry about it on 10-10ths. Also, if the throttle is applied too quickly, and more is asked of the tyre than it has at a given moment, it will lose traction too. But hard cornering and smooth throttle application can make a tyre give more grip than it would appear it can cope with.

To the original poster, you can't "add or take away load sensitivity". You're tyres have a sensitivity to load which you must work with. As you accelerate out of the corner you probably transfer more weight to the outside rear than the inside rear (though because there is a longitudinal acceleration both get some), and hence you will lose grip in terms of load sensitivity, and hence the slip angle will increase at the rear relative to the front. As long as you don't overdo it, this is what you want - the car ever-so-slightly oversteer biased (but not requiring opposite lock, or even neutral steering) on corner exit.

[Oldtony]

The other factor to be considered is the respective hieght of front and rear roll centres. This will effectively govern the wayin which changes in roll stiffness changes contact patch and thus adhesion of the tyres.
A rather old book, but a great basic guide is 'Racing and Sportscar Chassis Design' Mike Costin and David Phipps, which has all the formulas and a fairly easy to follow guide on how to apply them.

[adambrouillard]

Ok, maybe you can help me understand how the load transfer affects the front. I have always heard that transfering weight to the rear will cause understeer, but I'm not understanding how.

If you transfer weight off the front tires they will have less load, but because of load sensitivity shouldn't this actually create more "grip". I have always thought that the lighter the load of a tire the hire g potential it has.

I thought this was why you can tune balance by having more roll resistance at the front. The outside front gets more weight and thus lower g capability, the inside front gets less weight and thus more g capability, but the net result is a lower total g capability than the rear.

I have been trying to wrap by brain around this for 2 days now.

[tristancliffe]

Not quite.

A tyre will have generate more lateral force with more load. Always. The relationship isn't quite linear though, so doubling the load does not double the force. The lower the load the higher the coefficient of friction.

This is why a heavy car can't corner as fast as a light car - if it weighs twice as much it doesn't have twice as much grip, but still has to turn twice the mass. It also explains downforce - more tyre load, but without extra mass to turn.

However, tyre load sensitivity comes into play more when you look at an axle or both tyres on one side. Because the relationship between force and load isn't linear, it comes to pass that two tyres with, say, 100kg load each will produce MORE turning force than one tyre unloaded (in the air, say) and the other with 200kg.

For this reason load transfer is always bad, which is why everyone lowers their CoG and increases their track/wheelbase as much as they can get away with. Of course, other priorities like roll centre control, aerodynamics, camber control, stability etc all encroach on that and ensure that the best we can hope for is a compromise.

[adambrouillard]

I understand that concept, maybe I didn't explain what I was asking good enough.

Let's say a car is totally neutral 500 lbs on each corner with equal roll resistance front and rear.

You are cornering at 1 g which for this example will transfer 200 lbs.

So now each outside tire is supporting 600 and each inside is 400.

At this point the front tires are on the limit and the rears slightly behind due to different tire sizes. Now we begin to accelerate and have transfered 200 lbs from the front to the rear. The rears are now on the limit.

The outside front is 500 the inside is 300 the rears are 700 and 400.

My question is what has happened to the front? It has less weight so the lateral force is lower, but it also has to move less weight so the overall g capacity should be higher, correct?

Although I suppose since you are now going faster because you are accelerating, you may still have overcome the front tires.

[adambrouillard]

After thinking about it I think that actually is the answer. The acceleration would be the same affect as moving the cg rearward which would give the front more relative grip, but you will still be accelerating to the point that you either reach the front or rear grip maximum, and if you don't have enough power to overcome the rear tires, then you get (more understeer). Seems simple now.

So therefore as you accelerate more you shift more weight to the rear which would give the front more and more relative grip, which in turn means you need more roll resistance in the front in order to stay neutral depending upon the acceleration potential of your car.

So oval type course with not much acceleration you would run more even roll resistance. A course with lots of tight corners and big accleration zones would call for more front roll resistance to keep the car neutral under accleration.

Does this sound right?