The Mark Ortiz Automotive
Presented free of charge as a service
to the Motorsports Community
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.
ANOTHER PONY CAR SUSPENSION FROM THE '60'S
Last issue we took up the independent suspension on the later-version Corvair and the C2/C3 Corvette.
I have received e-mails from a number of correspondents regarding the form of independent suspension shown on pages 61 and 62 of the October 2007 Racecar Engineering, in Richard Nisley's article entitled "the Program Makers". This is an independent suspension using a reverse wishbone as one of the links, to control toe. The article states that the system shown was used on the GT-40, and a number of other cars. This idea was invented by Ford engineer Klaus Arning and patented by Ford in 1958. Correspondents have inquired what I think of the system, as shown in the magazine. One of them sent me a photo of the system shown on p.61, in place under a Mustang. I am sending that with this newsletter, as a separate attachment.
This suspension is described as a 4-link system. As we noted last month, it really takes five links to locate an upright, if by a link we mean a simple rod or tube with joints at the ends, loaded purely in tension and compression. To reduce the number of locating members to four, it is necessary for one of them to be designed to take bending loads. The Arning system uses what has come to be known as a reverse wishbone to accomplish this. The lower control arm cosists of two converging tubes, with two joints at their outboard ends and a single joint at the inboard end.
I haven't read the actual patent, but I gather that what is claimed in it is the combination of a reverse wishbone and two longitudinal links, which can be either leading or trailing. The illustrations show a version with the lower link trailing and the upper link leading, so that in side view the system looks something like a Watt linkage for a beam axle. The photo on p.61 shows an outboard brake. The drawing on p.62 shows an inboard brake.
I was surprised to learn that the reverse wishbone idea was patented by Ford, because it was first used on a race car by Lotus – but not before 1958. Lotus was using it in 1960, on both the Mk. 18 F1 car and the Mk. 19 sports racer. In fact, all rear-engined Lotuses used it, for at least ten years, and by 1962 it was pretty much universal in F1 and for rear-engined sports-racing cars. The earliest use by Ford was on the V4-engined Mustang I prototype, which as I recall was first shown in 1962. By the time Ford showed the public a running car with reverse wishbones on it, racers everywhere were building cars that way, and a number of small companies were building them for sale.
I don't know of any lawsuits seeking royalties from any of these small builders. Some of the cars used Ford engines. Others didn't. It would be interesting to know exactly who communicated what
to whom during that time. We can't necessarily assume that only one person came up with the idea. Sometimes when an idea appears in different places at about the same time, it really was one person's idea, and they communicated it. In other cases, the collective evolution of scientific thought leads totally independent individuals to similar conclusions nearly simultaneously.
However, all successful suspensions of this type have used two trailing links for longitudinal location, rather than the version with the leading upper link shown in the pictures. The system shown was used only on a prototype. It was never raced or mass-produced.
As a way of adapting independent rear suspension to a car originally designed for a live axle on leaf springs, the layout is appealing. It can anchor to the frame or unibody rails in approximately similar places. There are no upper trailing links intruding into the rear seat area.
However, for the layout to work as it should, the brakes have to be inboard. The reason why provides a good example of the difference between a case where torque acts through the suspension linkage and a case where the linkage only reacts thrust.
It is possible to arrange the geometry so the hub moves in a vertical straight line in side view. In that case, anti-squat under power is essentially zero, although it may go to a small positive or negative value with pitch. With inboard brakes, the anti-lift will likewise be zero, with small variations due to pitch.
If the links are parallel at static condition, the side view instant center (SVIC) is undefined in the x or longitudinal direction. However, it can be said to have a z or vertical location, at hub height. As the suspension compresses, the SVIC appears behind the wheel, at hub height. With further compression, it remains at hub height but rapidly migrates forward. A line drawn from the hub to the SVIC remains horizontal. But a line drawn from the contact patch center to the SVIC slopes downward below ground level toward the front of the car, and this slope increases rapidly as the suspension compresses.
The former line, from the hub, determines the anti-lift in braking when the brakes are inboard. The latter line, from the contact patch, determines the anti-lift when braking. So with inboard brakes, the anti-lift stays pretty much constant, but with outboard brakes it varies dramatically as the suspension moves. As the suspension compresses, it rapidly gains pro-lift. As it extends, the effect works the other way, and the suspension gains anti-lift. This means that when a bump compresses the suspension during braking, the force trying to jack it back up rapidly grows. As the suspension
extends during braking, a force trying to jack it back down rapidly grows. The changes in jacking force oppose suspension compliance both ways.
I generally take issue with those who suggest that longitudinal anti effects lock up the suspension during braking, because if the jacking force remains fairly constant as the suspension moves, variations in load due to bumps still move the suspension. The range within which the motion occurs shifts, but the motion is not resisted more strongly. However, this suspension presents us
with a case where the jacking forces vary in the same direction as the spring forces, as the suspension moves. Therefore, the total support force from spring and linkage combined varies more with displacement than it otherwise would, and the suspension truly does become less compliant.
This is also a reason not to use a Watt linkage on a beam axle for reacting brake torque. It is important to mount the caliper to the axle housing or use a separate brake floater instead. It shouldn't be on the birdcage.
Interestingly, if the upper link is trailing and the lower link is leading, the anti-lift with outboard brakes changes the opposite way. The jacking forces change oppositely to the spring forces, and tend to amplify suspension movements. Conceivably, that could be beneficial, at least in moderation.
The correspondent who sent me the photo says that the layout originally used a Jaguar XKE (Dana 44) rear end with the Jag inboard disc brakes, but the brakes were moved outboard during development.
REALLY QUICK STEERING
While driving a street-car I became interested in the question of why it is we have multiple turns lock to lock for the steering wheel function. This necessitates the driver having to remove his hands from the wheel for tighter turns. It can be an issue in racing as well, especially in rallying and some types of open wheel cars (returning to the pits or negotiating some chicanes). Karts and motorcycles have sub-one turn steering. Both are known to be controllable and in the case of a motorcycle (or any bicycle) it is impossible to conceive of a rideable system with more than a fraction of a single turn (I had fun imagining what a three turn lock to lock motorcycle steering system would be like!).
My question is, why not have single-turn steering for cars (that is, the steering wheel would turn approx 360 degrees lock to lock)? It would eliminate the problem of removing hands from the wheel in order to make some types of turn. It would also give the driver a far better chance of catching a car that had entered a slide. He would be able to command a faster response from the steering system.
The main concern relates to straight line driving at medium to high speed. In this situation the vehicle is very (too) sensitive to small steering inputs. Drivers perceived this as “twitchiness” or
excessive response on-center. In cornering the problem goes away, as lateral loads are introduced which make the “gain” of a fast ratio steering system work in favour of controllability. In that case the car is perceived as responsive but not excessively twitchy (the driver is working against transverse forces in the steering system which effectively “balance” the input from the driver and “damp out” the twitchiness or excessive response). So the trouble appears to exist only on-centre at speed.
I have an interest in installing a single turn system in a road car but can’t think of the best way to sort out the excessive on-centre response at speed. The power assist system does not appear to be much help as even disconnecting the pump has little effect on what occurs above 15 mph anyway (no assistance required in the medium speed on-centre cruising regime). It’s only at low speed you need the power assistance and at low speeds “twitchiness” is not present! What would you recommend?
Whether power steering is needed at speed depends on the car. Some cars require a lot of effort if the power steering quits. A two-ton car with one-turn steering and dead power assist could really be hard to control.
I drive full-size cars most of the time. Occasionally I get to drive a go-kart. When I get on the kart, I find it hard to be smooth, even when cornering. Practice definitely helps. A driver can get used to really quick steering, given time to acclimate.
One-turn steering without power assist is common on really light racing cars. Formula SAE cars very often have less than a turn. People would probably use kart-style non-geared steering in FSAE, if the rules didn't require geared steering. Assisted very quick steering is used in larger pure race cars, such as F1 and LMP cars. Sprint cars and midgets have very quick steering. In road cars, it's less common, but it has been done, notably by Citroen on the SM.
The SM had power-assisted centering, which varied with speed. This was done with a heart-shaped cam, with a hydraulically-loaded roller bearing against it. The hydraulic pressure was varied to control the centering force. The car had the peculiar habit of returning its steering to center by itself, even when stationary, as long as the engine was running. A correspondent who drove one in an autocross describes it as having horrendous understeer when used for this purpose. He surmises the designers were thinking more of straight-line superhighway cruising. Of course, this is mainly a handling issue, as opposed to a steering problem.
The combination I have found hardest to deal with is quick steering combined with a lot of lash (or play, or slop). Move the steering wheel a little, and nothing happens. Move it some more, nothing happens. Move it some more, and the car suddenly darts, and you need to correct the other way, and of course the same thing happens again. Even in this case, driver acclimation helps. You get used to moving the wheel rapidly until you start to feel resistance, and being delicate and precise from there, but it isn't as easy to do well as it is to describe. If you want to use very quick steering and have it driver-friendly, it's important to keep the slop to an absolute minimum.
It is possible to make steering gears that speed up or slow down as you get toward full lock. Nowadays, it is most common to make the steering speed up toward the ends, but for heavy cars with non-assisted steering, it was once common to make the steering slow down away from center, to reduce effort when parking. Even a fixed-ratio rack-and-pinion system gets faster toward the ends, due to the shortening of the moment arm length at the steering arms as the wheels steer.
It is also common nowadays for passenger cars, at least upscale ones, to have power assist that diminishes as speed rises, without going away altogether.