It’s no secret that not all cars are created equal in terms of handling characteristics. By handling we, of course, mean cornering, but it is so much more than that. Weight transfer, traction, braking, acceleration, and other vital defining traits are heavily influenced by the chassis. Massive global scale OEMs build their cars to a set of numbers, and those numbers often reflect safety concessions, comfort considerations, and a litany of other factors.
Tubular chassis kits and plans may exist for various applications, and this article is by no means a builder’s guide, but the foundational knowledge of what makes a car handle is fairly universal.
We spoke with seasoned ride and handling engineer David Thilenius, who among racing his own cars has worked in chassis development and testing for General Motors, Toyota, Hyundai, and Kia. Currently, he presides over Thilenius Group, his third party testing company which conducts engineering and subjective evaluation for OEMs and support manufacturers.
“I went to the Russel Racing School, and to GMI (General Motors Institute), an engineering co-op that at one time was owned by General Motors,” Thilenius explained. “It gave me an opportunity to be sponsored by the GM assembly plant in Van Nuys, (California) and I worked for GM throughout college.”
After a break at GM, Thilenius said, “I went to work for Toyota, and finally got into what I really wanted to do — vehicle development, as a ride and handling engineer. I worked for Hyundai-Kia doing ride and handling development for about seven years.”
Thilenius is no arm-chair bench racer, he has a history glittered with racing accomplishments from around the world — making him a keen advisor to help us navigate the complex world of chassis and suspension geometry.
“I’ve been racing since 1984, I started off in the School Series with Formula Mazda. Then, I switched over to Skip Barber and ran its Formula Ford series for years,” Thilenius said. “I basically jumped straight over from the School Series into the Firestone Firehawk series. I missed open-wheel cars so I got into club racing a Formula 2000 car.”
Thilenius has worked on independent projects with Shinoo Mapleton of Sector 111. Among these exoskeletal exotics, he did development work on the Drakan Spyder, and a geometry re-fit of the Ariel Atom to tailor some compliancy and forgiveness for rough Southern California tracks.
In the performance aftermarket we are much freer to devise vehicle running gear and geometry to a specialized application. Not bound by ease of manufacturing, mass appeal, or a boardroom of number crunchers — chassis geometry is one of the fundamental areas to understand when attacking your track car project or race car.
Maybe you have a gutted or rotted-out shell of a car that you wish to modernize, re-invigorate, and campaign on the track or autocross. This ambitious endeavor will involve fabricating a replacement base structure or chassis. If you want your pride and joy to handle better than a soap-box derby car, it’s worth learning a few basic concepts.
What Happens Under Acceleration?
As a car accelerates there are changes direction or velocity, and weight transfer occurs — loading and unloading different areas of the platform. We all know the sensation of a car leaning under cornering, diving under braking, or squatting under hard acceleration — these are all manifestations of weight transfer.
The shifting of weight about the normal origin or static center of gravity (CG), makes for a dynamic and synthetically altered CG. If you could corner balance a car while it was in motion experiencing weight transfer, you would see changing loads — side-to-side and fore-aft. While this moving target CG is fleeting, it is revealing. “In lateral weight transfer you are talking about roll-centers and fore-aft weight transfer happen the same way,” Thilenius said.
Types Of Weight Transfer
“There are three kinds of weight transfer in a car. First, weight transfer due to the unsprung mass, but there is very little you can do about it. The second type is called geometric weight transfer — that’s what anti’s are. Anti-squats and anti-dives give you geometric weight-transfer, and it happens through the control arms rather than through the compression of the springs. The third is what we call sprung-weight transfer — what we all tend to think of when the springs compress and the rear comes up or down,” Thilenius explained.
The way chassis and suspensions systems are designed has a direct relationship to how these dynamics are translated into ride and handling characteristics. While a luxury town car will be supple and compliant over the bumps it will not be engineered to provide snappy turn-in, or weight transfer to optimize traction under power. Conversely, a supercar is built to approximate race geometry with few concessions to prevent spilling the drinks.
Understanding what’s happening when you apply a control input — with feet or hands — is the first step to understanding why a car reacts the way it does under the aesthetic wraps.
Formula open-wheel cars leave little geometry to the imagination. Photo source: Thilenius Group
How a car steers is the first thing most people think about when we mention handling. The sensation of control, either bolstered or undercut by steering-feel, is an undeniably crucial part of the driver’s tactile interface with the car. Steering geometry goes much further than a quick-disconnect aftermarket wheel. Many characteristics of a vehicle’s steering are locked in, and would require huge amounts of fabrication to change, however, other adjustments are widely variable and can be altered by an alignment shop.
Sensations that we feel behind the wheel are easily misdiagnosed, as there are so many contributing factors and systems that determine the end driving sensation. For instance the ever-deplored torque-steer found in front-wheel drive cars may not be as much the fault of the OEM as the driving surface.
“Phantom steer feels exactly like torque steer — the car is being driven by whatever is the more loaded wheel. On a crowned road, different wheels are loaded more than others and may induce phantom steer — it feels like torque steer but it’s not,” Thilenius revealed. “One of the issues is long spindle length, which McPherson struts will give you. The torque around the length of that spindle is what gives you what feels like torque steer.”
The most basic measurements that influence steering are the alignment parameters. Most OEMs play it safe in the name of tire wear, liability, and forgiving driving characteristics. The first and most often questioned of these is camber. Camber is the inclination of the wheel in the X-axis when viewed from the frontal projection. Negative camber implies a wider track on the ground than on the top like an A-frame, and positive camber translates into a “V” stance with a narrower base.
Changing the camber at static height is not the end of the story for most cars, camber gain under compression, into the negative spectrum, is commonplace and desirable for handling and stability. Positive camber situations are inherently unstable, and promote the tire to tuck under and potentially role off the wheel bead.
Imagine the difference between standing on the inside edges of your sneakers (negative camber) with your ankles straight versus the outside edges of your sneakers (positive camber), if someone pushes you from the side, the opposite side of your foot will be the one to bear the load, and will pivot — flattening the sole to the ground for stability. The cambered tires of a car react the same way under a weight-transfer cornering load, rolling to support the shift.
Toe-In Versus Toe-Out
Next in the tool bag of those wishing to affect their alignment for better or worse is toe. Toe is the inward or outward direction the wheels point at a static wheels straight position. Toe-in is like a skier learning on the bunny slopes, pointing their toes inward at the front to create a safe, stable plow.
Most OEMs build cars to include some degree of toe-in because it provides reasonably forgiving, predictable and safe steering characteristics. Toe-in is a known understeer-generating effect — meaning a car with excessive toe-in will push through corners rather than turn-in sharply.
Toe-out is just the opposite, visually at wheels-straight, each of the tires will appear to be steering out to their respective side slightly. Toe-out is rarely seen in large amounts, but when it is, it is under droop where the suspension is unloaded and allowed to hang. Toe-out induces oversteer, making for twitchy and snappy handling.
“A trick with front wheel drive race cars, to get them to turn-in is to toe them out in the front — it gets that inside front tire to a high slip angle before the weight transfers off of it,” Thilenius offered.
Last in the alignment parameters is castor, and segways our discussion into the geometries you can not change with the turn of a wrench at an alignment shop. Caster refers to the inclination of the kingpin through which the steering knuckle or spindle pivots. In almost all applications this is expressed as a number in degrees — with positive degrees referring to a rearward inclination at the top, like bicycle forks, and negative referring to a forward inclination like a tiller.
“Large amounts of positive caster gives you negative camber with steer angle. You want to induce negative camber in a corner under bump. At GM we did it with some static negative camber and castor, at Mercedes they were doing it with zero camber and a lot of caster — that’s just a philosophical difference,” said Thilenius.
Caster affects a number of steering characteristics and feels that are hard to grasp until they are experienced first hand. A more assertive hands-off return-to center and steering feedback can be expected with an increase in caster, while decreasing caster will make a car tend to continue turning once you take your hands off the wheel.
With an understanding of toe change and camber gain under bump and droop, we can begin to discuss topics like bump steer. The simple fact is your wheels do not move in simple uni-directional path when a bump is hit, or a weight transfer is experienced. Lots of monkey motion is going on — toe is changing, camber is changing, the track width is changing, and all this translates into a tire contact patch as ambulatory as a the car as a whole.
Bump steer specifically refers to a phenomenon of toe change throughout the stroke of wheel travel. The most common pattern is an increase in toe-in under bump, and toe-out under droop. Bump steer can be drastic under the worst circumstances, but designed out in racing applications. The name bump steer may be a little misleading as it rarely acts like a steering pull, more often it upsets directional stability in a straight line.
Heavy bump or droop scenarios expose bump steer. Photo source: Thilenius Group
Picture this: you are hauling the mail down a fast straight towards a crest, take the infamous Flugplatz of the Nurburgring, for example. As the car clears the crest the front wheels unload — become airborne even. With a high bump steer geometry this means the wheels will likely be toed-out upon re-contacting the ground — inducing a very twitchy initial return to Earth.
“You like to have some separation in height between upper and lower control arms so that you have better camber curves. A consequence of a compressed space between the arms is a lot of lateral movement in the contact patch, and a lot of toe change as the suspension cycles up and down through it’s range of motion,” Thilenius illustrated.
“Normally if you want anything to do with bump steer it’s toe-in under bump at the rear, because that’s and understeer effect to the car, toe-out in bump is an oversteer effect. The thing that you always hear, that you never want — toe-out at bump at the rear, is called reverse kinematics,” he concluded.
A much more perceivable design of steering happens progressively from lock to lock — Ackerman angle — what allows the wheels to track arcs of different radiuses depending whether they are inside our outside of a corner.
It is easy to dismiss until you visualize what is happening. If you were to drive your car through wet paint, and then slowly test the turning radius making a tight turn through at least 360 degrees you would be left with a big doughnut painted on the ground. The inner circle traced by the wheel on the inside of the corner and the outside diameter by the other wheel.
If these two wheels where linked together for a completely linear steering relationship, the inside tire would always be scrubbing because it would not have enough inclination to roll through the arc. Without Ackerman understeer would run rampant, and turn-in would be hesitant.
“A normal production-car has maybe 20 or 30 percent Ackerman. I knew what 20-percent Ackerman felt like, and then I got to get into a car with 60 or 70 percent Ackerman,” Thilenius recalled. “The beauty of Ackerman is that it really makes the car turn-in well. If you think about corner entry with a high Ackerman car — as you start to turn-in that inside tire gets to a high steering angle, and therefore high tire slip-angle very quickly. On initial turn-in, that happens before the weight has transferred off of it — which tends to yank the nose into the corner.”
By engineering an exponential steering angle gain the inside tire will be correctly angled to attack it’s tight arc, and the outside to follow it’s looser arc. The amount of Ackerman built into a car is not a one-size fits all number and depending on the racing environment, different tendencies may become more desirable.
“The problem with huge amounts of Ackerman is the inside front gets to such a high slip angle that it then gives up. For slow corners the car turns-in fabulous, but then in fast corners it washes out and you get understeer,” Thilenius cautioned. Clearly it is possible to overdose on many of the racing geometry designs, taking them too far or out of context can cause more harm than good.
“On a race track environment like a street circuit such as Long Beach — a track with slow 90-degree corners — I would probably want to run as much Ackerman as I could physically put into a vehicle because I want it to turn in and I don’t have to worry about the mid-corner understeer. If I were to go to a fast open track like Willow Springs, I would run very little Ackerman. The last thing you’d want is push in mid-corner,” Thilenius said.
One last consideration in the steering design is scrub radius. The scrub radius refers to the amount of movement the tire contact patch must make in order to turn when steering input is applied. In a car with zero scrub radius the tire pivots around a central axis without dragging the tire forward or rearward.
To find what sort of scrub radius your car has — mark the center of the tire contact patch on the ground and then draw a line through the spindle king pin or other turning axis to the ground. If the two point fall in the same place you have a zero scrub radius, when the tire is outboard it is positive and inboard it is negative.
Changing the width and offset of wheels can drastically alter the built-in scrub radius and hurt or improve handling depending on the situation. Imagine if your wheels and tires were spaced hyperbolically wide, perhaps several feet. In order to turn the wheels, the outside edge of the tire would have to move a much greater distance than the inside due to their unequal distances from the point of rotation.
With modern CAD modeling, geometry can be worked out well before the prototyping begins. Photo source: Griggs Racing Products
After all the mounting positions are determined in a chassis about half of the geometry has been established. The remaining variables fall into line with the addition of the dynamic components — arms, links, knuckles, and other bolt-on parts. Here is where we most often find the aftermarket working their magic to appease the consumers’ handling requirements.
The angles, lengths, shapes, and mounting points of these extremities influence how the aforementioned weight-transfer we started this article with, is translated into movement or lack thereof. A chassis and suspension system is all designed around the concept of load paths, that is, when a force is applied at one point how does that force move throughout the system.
Sometimes these load paths lead to parts that absorb and dampen them like springs and shock absorbers, other times they remain rigid and ultimately dump their energy elsewhere. The cosmos dictates that conservation of energy means all energy has to go somewhere, in some cases that can mean back to ground like electricity looking for an origin, which is generally not in our favor.
Anti-Dive And Anti-Squat
The inclination of suspension members dictates what a car does when you step on the gas or stomp on the brakes. Usually the intuitive reactions are built-in to aid traction under accelerating and braking respectively — but we can circumnavigate these all together with the black magic of the anti. Well, it’s not really black magic, it’s just physics. Don’t be afraid if you slept through trigonometry and science class, this is geometry for gear-heads. A roll center is what determines the datum point around which all other suspension geometry moves.
“With the inclination you can define the instantaneous center of the suspension, and then the height of the instantaneous center relative to the height of the center of gravity determines how much anti-squat or anti-dive you have. If the instantaneous center is at the height of the CG then you have 100 percent anti-squat — when you go to throttle in that car, the springs will not compress. Weight transfer still happens to the rear but it doesn’t happen through the springs, it happens through the control arm and it happens instantaneously,” Thilenius emphasized.
Anti-dive and squat are carefully engineered geometries that prevent the weight transfer of a car from being routed through the springs and dampers of the suspension — what this translates to is a level ride height under braking or acceleration. At first glance this might sound like a cure-all, but beware of snake-oil salesmen fabricating racing chassis.
This may sound fantastic in the name of reducing body roll, and other handling virtues, but there are consequences to this effect, “If you have a lot of that weight transfer, and it happens instantaneously, you shock the contact-patch and it breaks the rear-end loose,” Thilenius explained.
Roll centers are intangible points in space about which the car is allowed to articulate with suspension movement — think of the roll center as an axis running from the front of the car to the rear. The roll center is found by drawing lines in space through the control arms and intersecting them with lines drawn from the tire contact patch to instantaneous center (the projected intersecting point of the top and bottom control arms).
By manipulating the location of the roll center in relationship to the car’s center of gravity profound handling characteristics are altered, this is a case where millimeters can make miles of change. The most common manipulative techniques to harness from roll centers are anti-dive and anti-squat.
The nature of cars is diverse, and their chassis or suspension designs as far reaching as the schools of through that engineered them. With these fundamental concepts we hope you can looks at the workings of a car with a more critical and discerning eye. There is no one perfect geometry or suspension layout, everything is a compromise — and for more race cars an evolution, that changes from track to track and race to race.
Diving into your suspension with an eye to completely rebuild is an endeavor to take cautiously. The OEMs may not always have ultimate handling and performance in mind but the amount of research, effort and careful engineering they put into their designs should not be discarded too readily. Widely recognized forms and designs like the Mustang II front end are common swaps for a reason, they work. Chassis and suspension geometry are easy to mess up, but if you insist on starting from scratch these basics should get you going.