The Mark Ortiz Automotive


August 2016

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.





The other day I was looking at ads for race cars and I saw an ad for a very competitive SCCA GT 3 car which I believe said it had a Winters solid axle and incorporated a “Lotus-link”.  Is this a type of four-link with side location unlike either a Panhard bar or watts-link?  Anyway, what is a Lotus-link configuration?  What is complete configuration, how does it work, and what are advantages or disadvantages to “regular” three and four link setups for a car with a solid/live axle?


I haven’t seen that exact term before, but I would assume this is the layout used on most versions of the Lotus 7.  There are two trailing links above the axle and an A-arm or wishbone below the axle, with a single pivot at the bottom of the diff housing and two widely spaced pickup points on the frame, typically below the forward pickup points for the trailing links.  The wishbone provides both longitudinal and lateral location.  The roll center is approximately at the pivot under the diff.


The layout is simple, provides good location, and is compatible with space frames.  The low roll center reduces lateral tire scrub on one wheel bumps and requires that the rear suspension have a higher share of the elastic roll resistance than it would if the roll center were higher.  This reduces the effects of driveshaft torque on wheel loads.


However, the layout is not compatible with strategies that compensate for driveshaft torque – at least not without changing things considerably.  I guess it would be possible to retain the lower wishbone and combine that with an offset torque arm and birdcages carrying the brake calipers at both ends, each with two trailing links.  Of course, that would only somewhat resemble the Lotus design and would be considerably more complex.


One problem with the design in the original Lotus application has been that there are very high point loads at the pivot under the diff.  With 1950’s tires and the modest engine power that the car originally had, the system holds up reasonably well, but when people add modern racing tires and a stout engine, the axle housing becomes prone to structural failure at that point. 



Partly to eliminate this problem, the last version of the Lotus 7, the substantially redesigned Mark IV, had a different rear suspension design.  However, the original design can be made to work if the parts are beefy enough.  Placing the wishbone closer to the ground and the trailing links up higher helps.  Two separate diagonal links can be substituted for the wishbone.


When considering purchasing a car with this type of suspension, it would be prudent to ask the seller whether there have been reliability problems with the system, especially if the car has a lot of engine, tire, and downforce.  A Winters axle will probably hold up to the demands, provided the bracketry is well designed.  Another thing to look out for is use of adjustable rod ends (Heim joints) on the wishbone, particularly for the pivot under the diff.  It’s bad practice to load the threads on these in bending.  It’s especially bad when just one of those threaded shanks has to resist all the lateral forces that way all by itself, and the axle has no lateral location if it breaks.





We have two letters from the questioner here.  I considered editing them into a single question but I decided to leave them separate.


First letter:


I have some thoughts relative to throttle on understeer during corner exit on a Porsche.  I would like your opinion on them. (Your July column/May newsletter)


My view is a lot of throttle off-on steer during cornering is due to roll angle change caused by the trailing arms being at different angles (power causing roll angle change).  In this regard roll stiffness increase would be helpful.


Also a readjustment of elastic roll stiffness to make the front and rear as equal as possible taking into account the front and rear weight ratio (perhaps not totally equal).  Then steady state balance brought back by adjusting the roll centers.  Perhaps there is not enough variance possible to do this though.


At any rate the main thrust of this argument is that the relative contribution of the elastic vs. r/c to the anti-roll affects the power on-off steer.


I had a Porsche 914/6 that was impossible to drive fast on a track due to corner exit power on understeer.  Increasing the roll resistance greatly solved the problem.


What is your view?





Second letter:


I recant all that monkey business about adjusting roll centers etc.  However I believe roll change with power is the culprit.  I believe the base problem with a Porsche is that the trailing arms are too short.   Can’t help it with this type of suspension.


Suppose one were to hook up a big lever to a body so a pure torque could be applied.  Would the car roll and jack about the conventional roll center or would it be different?  Suppose the springs were replaced with a Z bar and roll was resisted by torsion roll bars only and a torque was applied and the roll and jack noted.  Then the Z bar and the roll bars were removed and the springs were put back in but with the same roll resistance and the torque re-applied.  Would the motion be the same?


What I am getting at is: does the percent roll resistance between the springs and roll bar affect where the body rolls and jacks?  If it does, this would affect the angles of the trailing arms.


I forgot to mention that on my 914 I also added solid bushings.  I could do this on a 914 but I don’t think you can on a 911.  If I remember right there is no geometric center line between the two pivots. It is established by the elasticity of the bushings.  One might control understeer by varying the relative elasticity of the two bushings.


Incidentally, when my 914 was understeering under power, turning the steering wheel didn’t do much at all.  It was going to go where it wanted to go.


I leave this to you to figure all this out.


First, we should note that the original question in the May 2016 newsletter dealt with a Porsche 996, not a 911 or 914.  The 996 has a multi-link rear suspension, not the trailing arms of the earlier designs.  However, it is interesting to consider the effects of the earlier designs on power understeer.


My May 2013 newsletter is entirely about thrust roll, which is my term for the effect we’re discussing here.  Readers can still order this issue, for free.  I won’t recap all of it here, but I will note that short trailing arms produce thrust roll, as the questioner has recognized, when the car is in a rolled condition, and the effect does increase as the roll angle increases and as the trailing arms get shorter.  Also, the effect depends on the respective thrust forces at the two rear tires, and these vary depending on tire loadings and are also influenced by the properties of the differential/locker/spool that the car has and by the properties of the road surface that the two tires are on at a particular moment.


The 911 and 914 both use semi-trailing arm suspension.  The pivot axis in both cars is angled in plan view to give some camber change in ride, some camber recovery in roll, and a roll center a bit above ground level.  This differs from a pure trailing arm system, where the pivot axis goes straight across the car and is horizontal, giving no camber change in ride, no camber recovery in roll, and a roll center statically at ground level.


The geometry in the 911 and 914 is pretty similar, but the 914 uses coilovers and has two bushings for the trailing arm, while the 911 uses torsion bars and has one bushing similar to the 914 and uses the bushing for the outer end of the torsion bar as the other bushing.  There is a flex plate that works the torsion bar, as in the earlier swing axle designs.  Both the 911 and the 914 can use solid bushings.


Roll center height does not directly affect wedge change due to torque roll.  The front/rear distribution of elastic roll resistance does affect it, at least up to the point where the inside front wheel lifts.  Beyond that point, no change in diagonal percentage is possible, but further roll will adversely affect outside front wheel camber.


To entirely eliminate change to diagonal percentage change due to torque roll, we do not need the front and rear elastic roll resistance to be equal.  We need the rear to have all the elastic roll resistance – which of course is not possible, except for a drag car.  If we use a lower rear roll center, we can add rear elastic roll resistance.  Therefore, roll center height does have an influence, indirectly.


If we applied a roll moment to the car with a big lever, with no lateral force applied and the tire contact patches free to float, and the car somehow located by some other means, the car would roll independently of the suspension geometry, and the changes in wheel load would be independent of suspension geometry, but the tire contact patches would displace laterally with respect to the rest of the car in a manner related to geometric roll resistance.


Having two completely separate springing systems for roll and ride, versus conventional springs and an anti-roll bar, offers no advantage as long as everything is linear.  The tires don’t know the difference.  However, having a completely separate springing and damping system for roll does allow somewhat more freedom in tailoring rising/falling rate effects and damping characteristics for the two modes.