The Mark Ortiz Automotive
Reproduction for sale subject to restrictions. Please inquire for details.
Mark Ortiz Automotive is a chassis consulting service primarily serving oval track and road racers. This newsletter is a free service intended to benefit racers and enthusiasts by offering useful insights into chassis engineering and answers to questions. Readers may mail questions to: 155 Wankel Dr., Kannapolis, NC 28083-8200; submit questions by phone at 704-933-8876; or submit questions by e-mail to: firstname.lastname@example.org. Readers are invited to subscribe to this newsletter by e-mail. Just e-mail me and request to be added to the list.
WEIGHT TRANSFER FUNDAMENTALS
I have been confused with some basic weight transfer fundamentals.
1. Assuming 100% smooth surface (no curbs/no bumps) with a rigid enough chassis do I need to have suspension springs in my car?
2. Stiffer springs help to reduce weight transfer. With that logic weight transfer would be minimum with infinite spring rate. Isn't that good if you are assuming 100% smooth surface?
3. Now if I consider bumps, is there any theoretical approach to find correct spring rate/natural frequency?
On a smooth surface, it is definitely possible to run with no suspension other than tire and frame compliance. Every go kart in the world does exactly that. I even recall somebody long ago trying a full-size sports racing car like that.
This was circa 1964. I donít remember the name of the car. I think I read about it in Sports Car Graphic. It was built for the US Road Racing Championship, which was the predecessor of the Can Am series. It was a rear mid engine, V8 powered sports racing car. It was not successful.
In the early 1920ís, Bugatti tried just having the rear end rigid, and a stiffly sprung beam axle in front. That didnít work very well either.
The suspension serves to absorb road irregularities but it also has a second function. It controls the load transfer distribution between the front and rear: what percentage of the total load transfer (weight transfer) occurs at the front, and what percentage at the rear.
For vehicles such as cars, the overall stiffness of the suspension has only a slight effect on total load transfer. Lateral load transfer does not primarily result from roll. Most of it would occur even if there were no roll at all. Roll does result in a small lateral translation of the c.g., but this effect only becomes large enough to be of concern in vehicles where the c.g. height is large relative to the track width, as in trucks and buses.
The real reason for most of the load transfer is that the c.g. is above ground level and the tire contact patches are at ground level. When the tires generate lateral force (centripetal force) and accelerate the car toward the turn center, the carís mass exerts an equal and opposite inertial reaction force (centrifugal force) away from the turn center. The tire forces act at the ground plane. The reaction force acts at c.g. height. All these forces are horizontal with respect to the ground plane. The forces have parallel but offset lines of action, so they create a couple.
The couple tries to overturn the car toward the outside of the turn. The tires resist this and hold the car upright. The outside tires press harder against the ground and the inside ones less hard. For the car to be in equilibrium, the sum of the moments must be zero, just like the sum of the forces. The total load transfer between the right and left wheel pairs, times the track width, must equal the total lateral force, times the c.g. height.
For fairly low, wide vehicles, the total roll resistance for the front and rear combined has little effect on the total lateral load transfer. However, the relative magnitude of the roll resistance at the front and rear determines the share of the total load transfer that occurs at each end. It doesnít matter much how stiff the whole suspension system is, but it matters how stiff the front is compared to the rear. In this context, stiff springing does not reduce weight transfer; it increases it. If the front has ĺ of the total roll resistance, it gets ĺ of the total weight transfer, assuming equal track width at both ends.
Note that roll resistance has geometric and frictional components as well as elastic (from springs and anti-roll bars), so itís not just springs that determine apportionment of weight transfer.
Even with very stiff suspension, the relative stiffness front and rear can be used to control load transfer distribution. This makes suspension useful even in cases where it only helps a little in absorbing bumps.
There are ways to calculate what natural frequency youíll get with a particular configuration, but there are only general guidelines regarding what natural frequency you want. Itís always a compromise. Mainly, itís a compromise between the need to absorb road irregularities, and the need to keep the tires upright and keep the c.g. low. In many cases, aerodynamic considerations enter into the compromise as well.
WEIGHT TRANSFER WITH ANTI-ROLL BAR VERSUS SPRINGS
My car is a VW Polo 9N3 GTI
MY2006 with the same VW/Audi 20V 1.8 petrol turbo engine from the VW MK4
Rabbit/Golf GTI. I know they were not sold in the states but if you
think along the lines of the current model Ford Fiesta ST in terms of chassis
layout (front MacPherson strut, rear twist beam), weight and power then that
would be close.
I'm looking to set my car up for Australian east coast national park roads (think British B-roads but rougher). I have the rear end more or less where I want it. The stock height 25 N/mm (142 lb/in) springs have been replaced by slightly lower 22.5 N/mm (130lb/in) linear rate springs. The reasoning behind this was to drop the COG and rear roll centre. Due to the resultant reduction in rear geometric and elastic roll resistance, a 3 way adjustable RARB has been added to bring rear roll stiffness to a level slightly greater than before. It is now softer in ride but stiffer in roll than before. Predictably now that I can get the car to unload the inside rear tyre, the front (standard spindles with -0.5 degree of static camber) is not able to cope since it has inadequate roll stiffness to keep the outside front upright. In tight corners, improved turn in still moves into understeer and then wheelspin on exit.
My question therefore concerns the appropriate front end setup. From some quick measurements/calcs I estimate that the front OEM springs are about 25 N/mm. It runs a 20mm FARB. My lower control arm pivots are already parallel and the steering arms a few degrees up at the tie rod ends, so lowering is out of the question for fear of adversely affecting the front roll centre height, camber gain in roll and bump steer. I'm torn between upping the front spring rate or upping the FARB to increase front roll stiffness and curb the understeer. My indecision comes from not knowing which method will work better at achieving the front roll stiffness I'll need and which method will work better with the open diff. I am leaning toward increasing the size of the FARB to keep the car compliant in ride and to make tuning the dampers F:R easier. Any pitching and diving due to the standard springs can hopefully in large part be controlled when I install these Koni sport dampers. But how this will work with front end traction out of corners compared to stiffer springs, I don't know.
So my question is, when a FWD open diff car transitions to putting power down coming out of a tight corner, is it not more likely to lift and spin the inside front when fitted with a stiffer FARB as opposed to stiffer springs? That is, assuming equal front roll stiffness in each case, in a stiffer FARB car won't the weight transfer onto the outside front tyre be concurrently trying to lift the inside front, whereas in the stiffly sprung car wonít the spring in the more independent sprung strut still be trying to push the inside front into the ground?
Any help would be greatly appreciated by myself and many others in the VW Polo community here.
An anti-roll bar is just another kind of spring, but there is a lot of confusion about its action and effects. One often hears it said that if you achieve a given increment of roll stiffness with stiffer springs, the springs plant the inside tire, whereas an anti-roll bar giving the same roll resistance tries
to lift the inside wheel instead. This misconception seems to be especially prevalent in Australia and New Zealand for some reason, but itís pretty common elsewhere as well. I think I recall Carroll Smith saying something like that in one of his books.
No disrespect to Carroll Smith or anybody else, but there simply is no truth to the notion that one source of roll resistance produces a different change in tire loading than another. All roll resistance has to act through the tires, and it necessarily changes their loading when it does so. The tires donít know where the loading on them comes from. The roll resisting moment generated by any front or rear wheel pair is equal to the lateral load transfer at that pair times the track width, always. Not only is there no free lunch, but there arenít even lunches with different prices.
Iíve written of this in the past but it probably bears repeating: one origin of the idea that anti-roll bars transfer more weight than ride springs of comparable stiffness is that people get confused about what constitutes comparable stiffness. That confusion results from the fact that there are two ways to express displacement in the case of an anti-roll bar because it connects three things rather than two.
With a ride spring system, we have the sprung mass and the wheel. One inch of vertical motion of one relative to the other is an inch of suspension displacement Ė simple. But an anti-roll bar system connects the sprung mass and two wheels. So is an inch of displacement an inch of relative motion between one wheel and the other? Or is it one inch of motion between each wheel and the sprung mass, which is two inches of relative motion of one wheel versus the other? Is an inch of displacement an inch per wheel, or is it an inch per wheel pair?
If we test anti-roll bar rates the same way as ride torsion bar rates, by holding one end stationary and deflecting the other end and measuring the force, we get a rate in pounds per inch per wheel pair. Thatís half the rate that the bar gives in pounds per inch per wheel. So if we put on a bar thatís a hundred pounds per inch stiffer measured that way, and compare the effect to stiffening ride springs with the same motion ratio by a hundred pounds per inch, we will find that the bar change has a lot more effect on the carís handling than the spring change. But thatís just because we chose to measure and express the barís rate in pounds per inch per wheel pair rather than pounds per inch per wheel.
I advise using a generous amount of bar and correspondingly softer ride springing, when the goal is to control roll while still absorbing bumps, provided the car will not bottom or suffer adverse aero effects if large ride motions are permitted.
With the questionerís car, the rear bar will need to be stiffened along with the front one. To put power down on exit with a front wheel drive car, especially with an open diff, the inside rear wheel needs to be off the ground or at the point of impending lift even when we are feeding power in and releasing the car in terms of cornering force. That means the rear roll resistance needs to be considerably in excess of that required to pick up the inside rear wheel in steady-state cornering.