The Mark Ortiz Automotive


April 2013

Reproduction for free use permitted and encouraged.

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:  Readers are invited to subscribe to this newsletter by e-mail.  Just e-mail me and request to be added to the list.





Why is the rear roll center in production cars always higher than the front? 


 This is primarily a feel issue as the driver always wants the rear to catch up to the front.  However in an ideal situation shouldn't they be the same?


Actually, it isnít quite true that the rear roll center is always higher than the front in production cars.  This is so in the vast majority of cases, but not all.  The exceptions are mostly rear-engined.


Also, among the majority where the rear roll center is higher than the front, the amount by which itís higher varies a great deal.


One explanation frequently offered is that having the rear roll center higher than the front makes the car feel directionally stable, or inclined to understeer, in transient maneuvers.  The idea is that the car yaws out of the turn a bit as it rolls, because the front of the sprung structure displaces laterally a bit more than the rear with respect to the contact patches as the car rolls, and that makes the car feel stable rather than twitchy.


Another explanation one sometimes encounters is that geometric load transfer has more effect early in corner entry than other components of load transfer, and therefore a higher rear roll center frees the car up (moves it toward oversteer) on entry, making it feel more responsive.


Those both make a certain amount of sense, but they canít both be true at once.  A higher rear roll center canít be good because it makes the car looser on turn-in, and at the same time be good because it makes the car tighter on turn-in.


I think the real reasons are more prosaic, or more practical and historical than purely theoretical.




The biggest differences in rear roll center height are between independent rear suspensions and beam axle rear suspensions.  With independent suspension, jacking starts to become noticeable when the roll center gets much above four inches.  With a beam axle, thatís not a problem.  Also, with a beam axle it takes a bit of cleverness (and complexity) to get the roll center lower than about seven inches while maintaining adequate ground clearance for street use.


For racing, ground clearance can be reduced, and beam axles can have roll centers as low as three or four inches with simple designs, or a bit less if we get tricky.


So we can pretty much say that the usable range of roll center heights for independent suspension ends where the usable range for beam axles begins.


The same constraints apply at the front of the car, except that beam axles at the front are seldom seen on modern cars.  This is partly because beam axle front ends are more prone to various kinds of undesirable oscillatory and steering feedback effects than independent front ends, but really itís more just a matter of packaging.  The engine is in the way.


At the rear, there are also generally packaging advantages with independent suspension, but a beam axle can be accommodated, and it saves a lot of money.  It also generally will have more modest maintenance and repair requirements.  For these reasons, beam axles remain a popular choice for the rear.


Since the usable range of roll center heights for beam axles roughly begins where the range for independent suspension ends, any car with an independent front suspension and beam axle rear unavoidably has the rear roll center higher than the front, unless something very unusual is done at the rear.


In trucks, the engine is usually higher, and often there is room for a beam axle.  Even then, the axle usually has a bit of a drop in the middle, and the leaf springs or Panhard bar will necessarily be a bit lower at the front than they can be at the rear, for reasons of packaging.  A lower front roll center also reduces lateral tire scrub on one-wheel bumps.  That minimizes the aforementioned undesirable oscillatory and steering feedback effects.  With parallel leaf springs, having the springs a bit lower reduces spring wrap-up in braking.


Nowadays, most cars are front-engined and nose-heavy.  Those with rear-wheel drive seldom have less than 55% front weight.  Those with front-wheel drive seldom have less than 58%.  With rear drive, it is better for traction and handling to have at least 50% rear.  However, in most cases this requires an amount of engine setback that cannot be obtained without making the front of the car longer than it would otherwise need to be, or giving up engine accessibility and passenger space.  Also, there are advantages in ride quality if the car is nose-heavy and has the engine roughly between the front wheels.




The front and rear tires will generally be the same size.  Occasionally the rears will be bigger than the fronts, but the reverse tends to look a bit odd and is never seen.  To avoid excessive understeer,

the rear suspension needs to have more roll resistance than the front, at least relative to the weight it carries.  This can be obtained elastically Ė with springs and anti-roll bars Ė and/or geometrically, by using a high roll center.


If the rear suspension is independent, we may have the roll center a bit higher than the front, but we will have to use an anti-roll bar in most cases.  With a beam axle, we have a choice.  The higher we make the roll center, the less anti-roll bar weíll need.  We may be able to use none at all.  Having little elastic roll resistance at the rear makes the car soft overall in warp.  This reduces torsional loadings on the sprung structure on irregular surfaces.  That, in turn, allows the sprung structure to be less torsionally rigid, reducing weight.  With lots of elastic roll resistance at both ends of the car, we have to stiffen up the body/frame to avoid having cowl shake or torsional oscillation on bumpy roads, and to avoid having the car twist to the point where door operation is affected when the car is parked on a surface that causes a large warp displacement.


Having the car soft in warp also improves traction on uneven surfaces, especially with an open differential.


With a live axle, one disadvantage of having little rear elastic roll resistance, relative to the front, is that we get more torque wedge due to driveshaft torque, assuming we donít have rear suspension thatís designed to geometrically compensate for driveshaft torque.  For this reason, for road racing or high-performance street use, there is some advantage to having the rear roll center as low as possible, and using correspondingly more rear spring and bar.


Even with independent rear suspension, more often than not we see the rear roll center higher than the front, but not by much.  The difference between that and having both roll centers at equal height will not be very significant in terms of car behavior or feel.





I am trying to understand the operating principle of the F312 Dallara front suspension. Rather than attaching the pushrod to the lower control arm it is attached to the upright.
This would certainly reduce the loads on the lower bearing, but from a geometry perspective the only benefit I can see is to jack weight into the outside tyre when turning into the corner. This jacking effect would vary depending on castor angle, kingpin offset, wheel offset and the pushrod mounting position. Am I understanding this correctly? Or is there another reason for this system?


I think I first saw this in the late 1980ís, on a Formula Ford or Formula Continental.  Compared to having the pushrod attach to the lower control arm, it may or may not reduce the overall magnitude of the load on the lower ball joint or spherical bearing.  However, it eliminates axial loading of the


spherical joint, and thatís good.  Assuming the spherical is mounted ďflatĒ, so the bolt through the middle is roughly vertical, then if the spring acts through the lower spherical, the force holding the

car up acts axially on the ball in the spherical.  Sphericals are not very strong or wear-resistant in that direction.  They like to be loaded radially instead.  They can be loaded axially, but they have to be sized generously and replaced relatively often.  Large sphericals tend to have meager misalignment capacity compared to smaller sizes, and consequently may not allow sufficient suspension travel in some applications.  Long control arms help with this.


I have seen cars with the lower spherical mounted ďon edgeĒ instead, so the bolt is horizontal, running front to rear.  That makes the support load radial.  However, the joint will still see axial loading in braking.  This design also limits steering lock or travel, approximately to the misalignment capacity of the spherical.


Assuming the ball joints are somewhat inboard of the wheel centerplane, the lower control arm sees a tension load when the car is sitting still or running straight.  If the pushrod is attached to the upright near the lower ball joint, that tension load is increased.


In cornering, the load on the lower control arm can reverse, depending on the y and z position of the joint.  In the Dallara F312, the lower joint is very high: close to hub height.  This is due to the nose and the entire suspension linkage being high for aerodynamic reasons.  In such a case, having the pushrod attached to the upright probably does reduce loading on the lower joint in hard cornering, perhaps even keeping it from reversing.


The questioner is correct that there is an effect on steer jacking (suspension jacking as we steer, which changes wheel loads and also ride heights).  If the pushrod is ahead of the steering axis, the effect tends to roll the car into the turn (or reduce the usual outward roll from caster jacking), and add load to outside front and inside rear tires (i.e. wedge the car).  If the pushrod is behind the steering axis, the effect is reversed: it de-wedges the car, adding to the effect from caster jacking.


Or, it is possible to position the pushrod attachment point exactly on the steering axis, and have no such effect at all.  It would even be possible to position the pushrod mounting point inboard or outboard of the steering axis, and have the car jack the same direction, either up or down, at that corner in both directions of steer, similar to the effect we get from front-view steering axis inclination.  If equal on both sides of the car, such an effect does not change wheel loads as we steer, but it does induce a self-centering or de-centering force in the steering.


The nice thing about this is that with the pushrod attaching to the upright, we can to some degree separate steer jacking effects from the actual steering geometry.  This potentially offers the opportunity to better optimize the overall system for steering feel, wheel loading, and camber control together.