The Mark Ortiz Automotive


June 2017

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.





I recently encountered a rear-engined Renault that has centering springs built into its steering rack.  I never saw that before.  Why did they do this?  Is this a good idea?  A crutch for bad design?


I don’t think they’re a bad idea.  There are other ways to make the steering seek a vehicle centerline center, but they aren’t necessarily better.  The most common way is to use some front view steering axis inclination (SAI) combined with some positive front view steering offset (ISO term) or scrub radius (SAE term).  This causes both front wheels to gain positive camber with steer and lift the car at the lower ball joints – or more precisely, at the steering axis/ground plane intersection.  That induces a gravitational centering force in the steering: the car tries to follow gravity by straightening out the wheels.


This gets us the centering force with no added parts, but it also produces adverse camber change with steer on the outside front wheel.  This effect increases as we add SAI.  Steering offset is also a mixed blessing.  It makes longitudinal (x axis) forces feed back more through the steering.  That increases the driver’s ability to sense where the front wheels are because he/she can feel them going over bumps.  It also increases the driver’s ability to feel any pulsation in the front brakes, and increases the possibility of a sharp impact breaking the driver’s grip on the steering wheel or even injuring the driver’s hands.


In combination with caster, steering offset makes the car roll out of the turn with steer, and increases load on the inside front and outside rear tires.  In moderation this can be good.  In excess, it can make the car twitchy.  The optimum amount will depend on the car, the track, and driver preference.


At speed, caster combined with trail produces a centering force due to tire drag, and a force trying to steer the wheels out of a turn or down a lateral slope.  In moderation, this is good because it helps the driver sense lateral force at the front tires.  How much of this is desirable will vary with driver preference.



Caster combined with trail also produces a de-centering force at low speeds and when the car is stationary.  In extreme cases this will make the front wheels flop over to either side when the steering wheel is released, as often seen with dragsters.  In less extreme cases, at parking speeds the steering will have a normal feel at small steer angles but the force at the steering wheel will lighten and then reverse as we steer further from center.  Since most cars don’t do this noticeably, this will feel odd to a driver who isn’t used to it.  Centering springs can reduce or eliminate this effect.


I did a brief search for Renault Dauphine, Caravelle, R8, or R10 alignment specs and came up dry, but I think I remember reading back in the ‘60’s that they had around eight degrees of caster.  I drove a rented R8 in 1966.  I didn’t know it had centering springs.  I thought the steering feel was good: well weighted; appropriately geared; communicative, especially of lateral force; no odd quirks.  I also remember noticing that the front wheels tilted more than usual in the direction they were steered, indicating a lot of caster.


Rear-engined Renaults aren’t the only cars to use centering springs.  Corvairs have a torsional rubber bushing where the Pitman arm meets the relay rod.  That would give a little centering effect, in addition to some vibration isolation.  It’s common to see a pair of gas shocks on the tie rods of 4WD trucks.  Those mainly provide damping, but there’s also a little centering force from the gas springs.


With pull rod suspension, it is possible to arrange the geometry so the front end lifts as the wheels steer, with a lot less camber change than is created by SAI.  This is also possible with push rods, but not as easily.


The nice thing about centering springs is that they can easily be made as stiff or soft as desired, allowing the centering force to be tuned as needed without changes elsewhere in the system.  They allow SAI, steering offset, caster, and trail to be optimized for other design objectives, without any need to compromise to obtain centering force.  Then again, it’s usually possible to get very good steering without centering springs.  So they’re not about to become universal, but they shouldn’t be dismissed as a bad idea or a crutch either.





I’m designing a special and considering using DeDion rear suspension.  I know many successful cars used it in the ‘50’s and it would appear to have good camber properties.  However, I recently learned that Mercedes had a problem with theirs in their race cars of the 1930’s.  Apparently the cars had a tendency to “judder” while cornering, particularly on bumpy surfaces.  This reportedly caused Mercedes to revert to swing axles for their W196 in the ‘50’s.  However, Alfa went from swing axles to DeDion when developing the 159 from the 158.  Maserati used DeDion systems with great success.  What’s the story on this “judder” thing?  What causes it?  Is it an inherent problem in DeDion systems?




The phenomenon we call judder is basically wheel hop, most often during lateral acceleration (cornering).  It doesn’t necessarily happen just on bumpy surfaces.


DeDion suspension is like beam axle suspension, except that the final drive unit is mounted to the sprung structure as in an independent suspension, with jointed shafts going out to the wheels.  Compared to a live axle, there is a reduction in unsprung mass, and driveshaft torque does not react through the suspension.  The system gives 100% camber recovery in roll, if we disregard roll due to tire deflection.  Camber does not change with ride height.  Thus, camber does not change adversely as the car traverses humps and dips in a corner, or when the driver is compelled to lift off the throttle when cornering hard.


Judder in cornering can occur with beam axles, swing axles, DeDion suspension, or even independent systems other than swing axles.  It is sensitive to tire properties and grip level.  It is sensitive to compliance in the system.  But most of all it is sensitive to roll center height or lateral jacking coefficient.  I have seen it in swing axle cars, including a VW Squareback and a swing-spring Mk.3 Triumph Spitfire.  I have seen it in Chevrolet Chevelles with live axles on rubber-bushed triangulated four-link suspensions.


The Mercedes version of the DeDion had a DeDion tube with a swivel in the middle and a sliding block for lateral location that moved in a slot in the rear face of the final drive.  The tube assembly had to be able to twist freely because its ends were located by one simple trailing arm each.  The system couldn’t move in roll unless the tube could twist.  This geometry gave more than 100% camber recovery in roll: it caused the rear wheels to theoretically lean into the turn a bit, although this was not visible when the car was running.  The geometry also gave a roll center considerably above the height of the siding block, which was roughly at hub height.


The track was narrow by modern standards at only 55”, and the tires were very tall.  According to Denis Jenkinson’s book on the W125, rear tire section was always 7.00” but the rim diameter varied according to the circuit.  19’’, 22”, and 24” rim diameters were used.  Front rim diameter was always 22”.  With the middle choice of rear rim diameter, allowing half an inch of tire deflection, the height of the sliding block would have been nearly a foot and a half.  The roll center height would then have been around two feet!  A line drawn from the contact patch center to the roll center would have been at nearly a 45 degree angle to the ground.  I don’t know how well they were able to control clearances and deflections in the tube assembly, but there must have been considerable lateral compliance in those tall, narrow wire wheels.


So that’s why it juddered.  A DeDion system with a roll center no higher than a foot or so should be fine.  Auto Union, on similar tires and wheels, went from swing axles to DeDion for the D Type, but with a much lower roll center, and they are not reported to have had any such problem.


Mercedes, like so many other manufacturers, went to DeDion suspension to remedy the bad handling they’d had with swing axles, in their case on the W125’s immediate predecessor the W25.  The W125’s problem could possibly have been solved by just substituting four parallel links for the


trailing arms. That alone would have lowered the roll center to the height of the sliding block, and also have eliminated the need for the swivel joint in the tube assembly.  It would have been desirable to lower the roll center further, but it’s hard to say whether that would have been necessary to eliminate the judder.  In any event, the problem was clearly the result of detail misapplication of the DeDion design concept rather than the concept itself.


I think probably Mercedes’ reversion to swing axles for the 196 was driven more by marketing than by engineering; all their road cars had swing axles at that time.  But to make swing axles work at all, they had to adopt the “low pivot” version of the design also used later by Porsche.  Even then, the

cars were touchy to drive and required complex rear ride height control to minimize camber change as the fuel burned off.  According to Moss, the cars were not bad, but you had to be very smooth and never lift abruptly when cornering.