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.
LATE CORVAIR REAR SUSPENSION
I know it has been a while since you spent any effort analyzing the Corvair suspension (addressing the late model, fully independent only) but I wondered about your thoughts on the general layout and where improvements could be made. Most of the rulebooks now allow a lot more “deviations” from stock than they used to. I am building a late Yenko “Ringer” (looks like a Stinger) to do some open-track work. I can describe the suspension changes pretty easily: stiffer urethane bushings in the front end; a coil cut to lower it. The one variable I am still working on is the front mounting of the rear trailing arms. Stock had a rubber bushing with an auxiliary link part of the way back on the trailing arm, tied into an inboard location on the transmission mount. The vertical location of that inner pivot point determines the L/R movement of the trailing arm during roll. Chevy had at least two different mounts, attempting to add a bit of roll “Turn-in rear steer” probably to counteract the natural rear engine tail happiness of the Corvair. Several folks have just removed all the toe-control apparatus, replaced the rubber in the trailing arm with a Monoball, lowered the car, set the toe to some spec and gone racing. With fairly stiff anti-roll bars, the car still leans some in corners, but was pretty drivable on the road courses. Do you think the Monoball replacement, and the giving away of any additional toe change is a worthwhile tradeoff?
A picture of the Corvair suspension, scanned from the GM shop manual, accompanies this newsletter. This happens to be a car I know well, having owned one since 1976 (with a Crown V8 conversion since 1978).
For readers less familiar with the car, the Corvair is a rear-engined sedan (though with a fairly cramped rear seating area), produced by Chevrolet from 1960 through 1969. The engine is an air-cooled flat 6, hanging out behind the rear axle as in a Porsche 911. This puts around 62% of the weight on the rear wheels, varying somewhat with equipment and fuel load, and gives the car a lot of yaw inertia.
There are two basic designs of Corvair: early (1960-1964) and late (1965-1969). The early and late models have dramatically different styling. I call the early version the sardine-can body, for its
somewhat brick-like shape with protruding lip all around the beltline. I call the later model the Coke-bottle body, for its curvacious fender line. There are also pickup truck and van models, called forward-control vehicles, similar in layout to VW transporter models, which were based on the early platform and were made from 1960 through 1965.
The early platform had swing axle rear suspension, and earned itself a bad reputation for not only having snap oversteer at the limit like a VW or Porsche, but also having an unusually compliant spring and tire combination from the factory that allowed the outside rear wheel rim to actually contact the road in extreme conditions and sometimes hook an irregularity and flip the car. The early models can be improved dramatically with lower ride height, stiffer springing, and good tires. Really, they like a lot of ride stiffness, especially at the rear. The factory attempted to add ride rate at the rear, and roll rate at the front, in 1964. At the rear, they added a transverse leaf spring acting only in ride, similar to the aftermarket EMPI Camber Compensator, and they added an anti-roll bar at the front. Porsche also used a swinging transverse leaf on the last versions of the 356, and Mercedes-Benz had a coil spring version of the same concept on some swing axle rear ends, as early as the 1930's. The "swing spring" on late-model Triumph Spitfires is another version of a swing axle system with a higher wheel rate in ride than in roll. "Zero roll" rear suspensions on swing-axle Formula Vees take the concept to its logical conclusion: no elastic roll resistance at all; springing that acts in ride only.
In any case, the early models have a handling problem. It can be fixed, at least somewhat and for certain purposes. Set the early cars up low and stiff, and they work much better than you'd expect, especially for low-speed events like autocross and hill climbing that favor a loose (oversteering) car. This is a case of any suspension working if you don't let it.
By 1963, when people started suing them, GM realized there was a problem, and set about remedying it for the next version of the car. The redesigned rear suspension was similar to the system introduced in 1963 on the new Corvette Sting Ray, and found on all C2 and C3 Vettes. In the transverse plane, it's a short-and-long-arm (SLA) system, using the halfshaft or driveshaft as the upper arm. In side view, it is a trailing arm system. The "torque control arm" is stiff in bending both vertically and horizontally. It does actually react torque in braking, although not under power. It also locates the wheel longitudinally and for toe.
In the Corvette version, springing is by a non-swinging transverse leaf spring, and the toe-controlling "front strut rod" shown in the Corvair suspension illustration is absent. The Corvette has a fairly serious deflection steer problem under power, due to compliance in the bushing at the front of the torque control arm. This bushing is much stiffer than in the Corvair, but it is rubber and it has to accommodate rotation in three axes. Fortunately, the wheels toe in under power rather than out, but the deflection is greater than one would wish, especially when the car is fitted with wide, sticky tires and large-offset wheels. Racers running C2 and C3 Corvettes eliminate the bushing deflection by substituting spherical joints (monoballs) for the bushings. The bump steer properties are not perfect either way, but they aren't horrible and deflection steer is largely eliminated. People racing late-model Corvairs often do the same thing, as the questioner notes.
The Yenko Stinger was a mildly modified Corvair, with the back seat removed and fiberglass C-pillar extensions to visibly differentiate it from a stock Corvair, and make it look more like a coupe and less like a sedan. The reason people road raced these and not regular Corvairs had to do with what SCCA was willing to recognize as a sports car. If a car had a back seat, that a small adult could squeeze into at all, it was considered a sedan. Sports cars didn't have back seats. If the owner removed a sedan's back seat, that didn't make it a sports car. However, if the car came from the manufacturer that way, then it was a sports car. Yenko was the manufacturer of record for a Stinger, rather than GM, and they took the back seat out, so that made the car a sports car. Shelby did the same thing with the Mustang when creating the original GT-350.
There weren't very many Stingers made, and they have become scarce with the passing years, so now SCCA is allowing people to make reproduction ones out of old Corvairs. That's what a "Ringer" is.
As to what we might do if the rules are liberalized, a great deal depends on exactly what those liberalized rules allow, and also what the budget allows.
Really, the biggest potential improvement isn't a suspension modification at all. The thing the Corvair needs the most, that I see most people missing, is bigger tires on the rear. The engineers evidently knew this, because they left extra room in the rear wheel wells. They couldn't actually put larger tires on the rear from the factory, because GM had a corporate policy that all cars had to have one spare tire that would work on all four corners. But there was no rule that they couldn't leave room for enterprising owners to install larger rear tires after purchase, so they did. Why Yenko didn't seize the opportunity has always mystified me. My recollection from road test reports back in the day was that street Stingers had the same size tires at both ends. All the Stingers and Ringers I've seen race have wider tires than stock, but the rears are the same size as the fronts. Perhaps this has to do with Yenkos being sold through Chevy dealers, and SCCA not allowing unequal size tires for racing if the street model didn't have them. At any rate, even when the car is set up so the inside front wheel lifts, it has visible limit oversteer. This works okay for autocross, but for full-scale road courses the car is in crying need of more rear rubber.
The Corvair came with really tiny wheels, I think 13 x 5˝. Mine already had 14 x 7 chromies installed by the previous owner, with 185/70 tires. With zero offset (4" back space), these bolt right on at both ends of the car. I left those on the front, and tried 15 x 8 Corvette wheels at the rear, with 215/60 tires. This combination also bolts right on, at the rear only, clears the fenders in all conditions, and absolutely transforms the car. Camaro Rally wheels on the front and Corvette Rally wheels on the rear make a readily available, Chevy-looking package.
I still have antique aftermarket wheels on my car: heavy but nice-looking Shelby Vector cast aluminum, 14 x 7 on the front with 195/55 tires, 15 x 8˝ on the rear with 225/50 tires. These rears only clear the fenders if I run about 3 degrees negative camber, and more ride height than you'd want
for a track car. My wheels have 4" back space front and rear. There looks to be at least an inch of room available on the inboard side of the rim and tire, both front and rear. With more modern
wheels, it should be possible to get at least 9" of rim width at the rear, and 8" at the front, without modifying the bodywork.
What diameter those wheels and tires should be depends on a number of factors. Some of these are straightforward packaging issues. Others have to do with the limitations of the suspension design.
To locate an upright using only tension/compression links, we must use five links. For a rear suspension, this normally means two longitudinal links and three transverse ones, or something approximating that. We can use fewer parts than that, but then some of them have to resist bending loads. The Corvair suspension has the three transverse links, with the halfshaft as one of them, but the designers simplified the longitudinal location by using a rigid beam, pivoted about a single point, anchored rigidly to the hub carrier, and loaded in bending.
This compromises the side-view geometry, at least when using outboard brakes. Because brake torque reacts through the arm but drive torque does not, the system's longitudinal anti's are dramatically different for lift under braking and squat under power. We would like to have moderate anti-squat under power, and moderate anti-lift under braking. With this design, and outboard brakes, that is not possible. The only way to keep the anti-lift halfway within reason is to have the pivot somewhat below the hub. However, that gives us pro-squat under power, so we can't lower the pivot too much. Making the arm longer improves the compromise, but the arm starts to get heavy, and packaging constraints limit us.
If we make the wheel taller, and lower the suspension to keep a given ride height, the anti-lift stays about the same, but the pro-squat increases, which is not good. We can live with a lot of anti-lift if our brake bias is such that we don't approach rear wheel lockup. We can live with squat under power if we have enough ride rate and available compression travel.
The transverse or front-view geometry presents us with a different picture. Here, the designers saved parts by using the driveshaft as a suspension link. This saves both cost and unsprung weight, but it entails geometric compromises. If the U-joint angles are to be kept moderate, the halfshaft needs to be fairly close to horizontal. If the halfshaft is horizontal, and we want 50% camber recovery in roll, the front view instant center for each rear wheel is at the center of the opposite wheel, and the roll center height is around half a loaded tire radius, or about 6 inches. That makes the force line angle of elevation roughly eleven degrees. That's enough to produce visible jacking if the tires are sticky.
The geometry gets better as we raise the hub or lower the car. Unfortunately, the U-joint angles rapidly increase. Up to a point, we can live with that, especially if power is modest, but there is a penalty in frictional losses even if we can keep the joints alive.
The C2 Corvette had front-view geometry much like the Corvair. With the skinny tires that car had, the jacking was not too noticeable, but when people started adding rubber, it became a problem. For the C3, Chevrolet lowered the inboard strut rod pivots. This increased the front view swing arm
length, lowered the roll center, and reduced jacking. Crown offered a strut rod mounting bracket for the Corvair that does the same thing.
The problem then is that the suspension loses camber recovery in roll. This is usually addressed by having lots of roll stiffness. Camber change in ride is reduced, which is good in a straight line.
As long as we keep the halfshafts as the upper control arms, we face a three-way compromise among U-joint angularity, camber recovery, and geometric anti-roll. We can't optimize all three at once.
A kit for improving C2's and C3's is still sold by Guldstrand Motorsport (see at http://www.guldstrand.com/scripts/prodView.asp?idproduct=80). It uses the original Vette brake and hub carrier, and adds a bolt-on upright, two transverse lower links, and two parallel trailing links, all with Heim joints. The halfshafts are still used as suspension links, and the lower links are nearly horizontal, meaning the compromise is weighted in favor of roll center height, at the expense of camber recovery, as in the stock C3 layout. The trailing links run slightly uphill toward the front, giving moderate anti-lift and anti-squat.
It would be possible to make yourself a similar setup for the Corvair – perhaps even use a Guldstrand kit with minor modifications.
Better yet would be a suspension that doesn't use the halfshafts as suspension members. We would like a front-view instant center around 70 inches from the wheel, and five inches above the ground. This requires the upper link to slope down toward the middle of the car, more steeply than the halfshaft can slope and still transmit power. The halfshaft then has to use splines or tripod joints, so it can accommodate plunge.
Such modifications should be legal anywhere that a Guldstrand suspension is legal on a Vette.
While we're considering cost-no-object modifications, how about scrapping the Saginaw 4-speed and using a Hewland, Porsche, or ZF transaxle? Once you're redesigning the suspension and halfshafts, it's tempting to throw out everything from the flywheel to the crossmember and just start over.
Of course, that's expensive, and a cost-no-object Corvair-powered Corvair really is an exercise in fanaticism. It's probably no crazier than a high-dollar Camaro or Impala, though. People build those, and in my opinion the Vair has nicer lines. I don't have a very high opinion of the stock engine, but it does have a low c.g., and it makes a nice noise. For those of us who have more limited means, the stock hardware really is not so bad, and it is reasonable to ask what is the best low-buck approach.
I like keeping the toe-control link and the big, soft bushing at the front of the torque arm. It makes sense to take compliance out of the toe link. You can make one with rod ends, or you can use the stock one, with urethane or nylon grommets instead of rubber. The grommets are ordinary anti-roll bar drop link grommets. To use four of them per link, you may have to cut them down a bit, or you may not be able to start the nuts – or you may be able to start them but not get safe thread engagement. The stock rubber grommets squash down a lot when you tighten the nuts, and they stay squashed if they've been in for a while. If you use stock grommets on the inner side of the torque arm and outer side of the crossmember bracket, and hard ones on the outer ends of the link, everything goes together with no cutting of anything. You get a link that allows compliance toe-in more easily than compliance toe-out.
When you lower the inboard end of the link, the wheel toes in as the suspension compresses. This provides roll understeer. To some extent, the understeer this creates isn't real at the contact patches, or at least the effect is smaller than the driver thinks based on required steering wheel input. However, there is some actual effect even in terms of slip angles, at least as long as we're applying power, which we usually are when cornering hard in an oversteering car. When the rear wheels are aimed more toward the inside of the turn, the car's c.g. is further toward the outside of the turn with respect to the rear wheels' lines of thrust, and that actually does add understeer.
Having the wheels steer in any way with suspension motion is at best a mixed blessing. One thing I've noticed when running ample roll understeer is that you get heave oversteer when you go over a crest while cornering hard. The rear steps out as it comes up on the suspension, then comes back in as the car settles back down. This mostly just feels disconcerting, but it does require a steering correction by the driver, and if you're really on the limit, it could cause you to lose the car.