When most of us think of new technology related to race car dynamics, we may immediately think of the major innovations coming from IndyCars and Formula 1. This is a natural conclusion, but because of recent developments, the true story is very interesting. In that particular field of science, stock car racing today may well be at the head of the class.
The evolution of the knowledge of vehicle dynamics and its history runs directly parallel to the development of passenger cars and trucks for several understandable reasons. First and foremost, the big automakers had the funds to finance extensive and expensive research. Also, the need existed to develop better suspension systems for ride comfort as well as drivability. Along with these two driving forces, there also existed the element of competition to develop more advanced cars than did other automakers in order to enhance sales. The "win on Sunday, sell on Monday" attitude has not been restricted to American automakers. Companies such as Ferrari, BMW, Toyota, Honda, and others hold the same to be true. Racing provides a unique venue to show the superior quality and advanced mechanical systems automakers have developed. Some consumers relate to and buy the cars associated with winning automakers.
Early pioneers of stock car dynamic research include individuals such as Maurice Olley and his group whose work represented much of the early progress that had been made in vehicle dynamic research. Among many other notable accomplishments, his work in the early '30s led the industry to adopt the double A-arm suspension system, or SLA (Short Long Arm) suspensions.
In 1952, Bob Schilling, head of the mechanical engineering department in the General Motors Research Laboratory Division, and his group met with a group of aircraft engineers, including Bill and Doug Milliken. The aero engineers were then contracted by GM to attempt to apply techniques that had been used in aircraft design to the study of land vehicle dynamics.
A compilation of that work, as well as other research, is contained in a book written by the Millikens and published by SAE (Society of Automotive Engineers) titled Race Car Vehicle Dynamics.
Competitive stock cars have usually been thought to be the "soap box derby" in all of auto racing. In several leading books on the subject of race car dynamics, stock cars are often referred to as a "particular case" and referenced by their unpredictable nature associated with handling.
It is fairly easy to adjust the handling on an open-wheel formula car with wings in the front and rear. A little more downforce at one end will effectively cure a tight ("understeer") or loose ("oversteer") condition. With the stock car, we do not have such options. The limitations and restrictions imposed by the "stock" chassis have, by their own admissions, confused many race engineers.
As the designs of exotic race cars developed, the engineers began to incorporate aero downforce into the structures of the cars. They added airfoils (wings) to the front and rear of the cars to enhance traction and to facilitate adjustments to the handling balance of the car.
Development related to chassis setup balance tended to get lost in the process. As stock car racing grew in numbers and the designs became less "stock" and more complex, racers renewed the quest for more complete information related to chassis dynamics.
During the period between the late '40s and the early '90s, no one had completely developed a way to predict how a stock car would handle. Consequently, no one was able to adjust the suspension components to attain that perfectly balanced setup. We basically knew that if we could ultimately predict the distribution of weight on the tires when the car was executing the turn, we would know how the car would handle. Historically, the predominant method used by teams to develop their setups was trial and error.
To better understand this, we need only to listen to the teams in the top three racing series today. In descending order (related to money spent for each type of race team), we quickly recognize the top three: Formula 1 is far and away No. 1, IndyCar/Cart racing is second, and NASCAR Winston Cup is third.
A casual observer would, based on the hype and subterfuge presented by the media, assume the technology existed that could accurately predict the handling characteristics of the vehicles, thereby resulting in perfectly handling race cars. When we listen to the drivers and crews, we hear a different story. In my opinion after careful observation, the Ferrari cars dominated Formula 1 (F1) in the year 2002 because they outhandled the competition. There were many races where they were admittedly underpowered, but still turned faster lap times, which comes from a better handling package. Even Michael Schumacher will tell you that he can only drive the car as fast as it wants to go.
It is the same story for American-based formula cars and Winston Cup teams. How many times have we heard a Winston Cup driver complain of a tight, loose, or unpredictable car? It seems to be hit or miss, so whether you're Schumacher or Tony Stewart, at each race you hope the crew has guessed correctly.
What has been refined in most top racing series is the hit-or-miss art of trial and error. Advanced measuring systems used today not only record movements, pressures, and temperatures, but also the forces exerted on components. These systems are useful and necessary tools for the modern-day chassis tuner and developer. As teams compile and study this information, the fact still remains that they tend to react to, and not necessarily predict, the handling nature of their cars.
Technology has evolved based on the ability to predict the handling characteristics of a car. This is a continuation of the work that early vehicle dynamics pioneers such as Olley started. Without their efforts, none of what comes next could possibly exist.
Developing and refining this advancement in technology was a fairly long process and involved many persons past and present, but the "who" is not nearly as important as the "how" and "why." The answers to those questions begin with a short study of the history of the development of the automobile.
Long before the car, we had horses and wagons. Transportation technology took a quantum leap with the invention of the locomotive trains. The first steam-powered locomotive was built in 1804 in Wales and was used in mining operations. As the design of trains advanced toward the late 1800s, so did our understanding of powered vehicles. It was inevitable that the automobile would become a vital mode of transportation.
As the auto became more popular, its chassis design became more advanced. The 1903 Ford Model A, the first production automobile, was built much like wagons with leaf-spring suspensions, similar to the stagecoaches of the early days. Power was transferred to the rear wheels via a chain and sprockets. The 1909 Model T incorporated a transmission and driveshaft for the first time, a system still in use today.
The front suspension of this car used one-piece spindles attached to the ends of a transverse (sideways to the centerline) leaf spring and a drag-link steering system, similar to that used today in many American production and racing stock cars.
Engineers working for the automakers developed the double A-arm front suspension system using coil springs. That system is the primary system used in stock car racing today.
So, the vehicle for which we are trying to predict the handling characteristics has a stiff chassis (further strengthened by a rollcage and supports at the front and rear), with a double A-arm, coil spring, front suspension, and a solid-axle rear suspension with either coil or leaf springs.
The front and rear suspension systems of these types of cars are very different from one another, both in appearance and in the way they react to the cornering forces that stock cars encounter. Therefore, we must treat each system independently and try to determine what each one wants to do as the car turns. If we can predict exactly what each desires to do, then we can match the "desires" of both to help create a truly balanced setup.
Modern Chassis Technology
Let's go back for a minute and see how chassis technology was progressing through the '50s to the early '90s. Engineers either working for or contracted by the automakers, with a heavy contribution from the General Motors group, undertook much of the research.
The primary thread of their analysis of vehicle dynamics involved a model of a vehicle that treated the body and frame as a single unit with a single center of gravity for the sprung mass. The roll centers at the front and rear were connected by an invisible line or axis, and a right-angle line between the center of gravity (CG) and the "roll" axis was the vehicle's moment arm.
A moment arm is much like a prybar or shovel handle. The CG is equivalent to the end of the bar we hold on to, and the roll centers and axis are the opposite end, which is the object we are trying to move. The longer the bar (moment arm), the more leverage we have and the easier we can move the object or, in this case, roll the car.
In the roll axis thread of technology, each end of the car was calculated to have a given "roll resistance" percentage based on spring rates and other pertinent information. In subsequent skid-pad testing, a neutral handling car was found to have significantly different roll couple numbers at the front and rear.
The handling characteristics of the car could be altered by changing the "couple" at each end, but the handling still could not be predicted. The cars were dialed in through trial-and-error methods. The "roll couple distribution" thread of vehicle dynamics analysis did not present a complete model that would by itself accurately predict a stock car's handling performance.
Because technology is, for the most part, universally shared among race car engineers around the world, we can be fairly sure that up until the early '90s, there did not exist a method that could be used to accurately predict the handling of either an F1 car or a stock car. In my personal discussions with motorsports engineers, I concluded that a definitive method and associated software did not exist at that time to accurately predict a stock car's handling characteristics in advance of going to a test or to a race.
In the early '90s, a method was developed that represented an advancement related to vehicle dynamics modeling. It involved treating the vehicle as two separate masses, each with its own separate suspension system.
This method made a lot of practical sense because, in a stock car, we have two "axles," each supporting the weight of that end of the car, and each resisting the lateral forces created by cornering. In physics, we are taught that everything is fluid, or somewhat flexible, and dividing the car into separate halves allows us to analyze each one independently to determine its desire. A review of early notes kept by a select few of the top stock car crew chiefs of the '80s and early '90s indicates that even then they were thinking in the context of a balanced setup. The goal was to arrange the setup so the two ends of the car would be balanced in each one's efforts, thereby providing a perfectly balanced handling car.
In 1994, a method was developed and algorithms (equations and math calculations) were written that would provide answers needed to achieve that goal. Weights and measurements from the car and racetrack were entered into a computer program, and through a series of calculations, a prediction for what each of the two suspension systems desired to do was concluded. Although the calculations were very complex, the two answers were presented as simple roll angles.
By treating each end of the car as an independent system removed from the other, we could now accurately calculate and predict to which angle each wanted to roll. It was discovered that the key to developing a balanced setup was to match the desires (roll angles) of each end. So, if the two roll angles are the same, we will have a truly balanced setup.
There are several critical reasons why a balanced setup is essential to optimum chassis performance. First, we can accurately predict weight transfer if the setup in the car is balanced. An unbalanced car redistributes the weight on the four tires in a very unpredictable way. If we cannot determine the desires of each end of the car, balanced or not, we then cannot accurately predict the exact amount of weight transfer and ultimately how much weight ends up on each tire at mid-turn.
Second, we will have less (almost non-existent) chassis flex with a balanced setup. Compliance, or flexing of the chassis, cannot occur if we remove the forces that cause this to happen. In 1995, a race engineer asked me what I thought of compliance. I asked if he meant chassis flexing between the axles, and he replied yes. I told him that if we could remove the forces that ultimately try to twist the chassis, we could mostly eliminate chassis flex.
Last and most important, we need to have two sets of tires that are each doing equal amounts of work at mid-turn. The pairing depends on the static weight distribution. In stock car racing, we will almost never have all four tires doing equal work (having equal weight on each tire) under most current left-side weight rules. If we can set up the car to have equal working pairs of tires after the weight transfers in the corners, then we will have a truly balanced setup.
If you carefully watch an F1 or a Winston Cup car going through a turn, you can judge what degree of balance exists. A car that is twitchy or rocks as it turns is not very well balanced. The handling may be neutral for the time being, but that does not mean both ends of the car are in sync. A car that loses a lot of speed on a particular run on the same set of tires is usually not very well balanced. The primary cause of the loss of speed due to an unbalanced setup is that the car's handling balance becomes tight or loose.
Unbalanced setups abuse one or more tires. It is this abuse that causes the tires to wear or heat up excessively. Balanced setups distribute the weight and workload evenly on each pair of tires.
The advancements that have and will come from this new approach to chassis setup will benefit all of racing. The fact that it came as a result of the unique character of the design of the stock car puts our sport right in the thick of vehicle dynamics development. Like most major technological advancements, a specific need came first, and then, through the efforts of experts in the specific field of study, came a solution.
In the next installments, we will learn about the specific dynamics of each suspension system and how we can measure and adjust each to find that perfectly balanced setup.
( From Private Model Collection...
(From Private Model Collection Of Kim Murphy)
The driver and crew chief...
The driver and crew chief need to communicate about setup. If someone guessed wrong, there's often little time to get it right.
Harry Hyde was a crew chief...
Harry Hyde was a crew chief who tried to balance his cars. He had his own method worked out and found much success in Winston Cup racing.
The typical stock car has...
The typical stock car has a front suspension system using double A-arms. These arms are attached to a spindle and the controlling dynamic point is called the "roll center." MacPherson strut front suspensions, a system with a roll center point similar to the double A-arm system, are being used more and more in stock car racing.
Roll Axis--roll couple distribution...
Roll Axis--roll couple distribution technology was primarily developed for symmetrical suspension systems such as those found in production automobiles. This method involves calculating the resistance to roll of each suspension system. It does not take into account all of the dynamic effects associated with a race vehicle.
The Roll Angle Analysis Method...
The Roll Angle Analysis Method allows us to predict how each end of the car wants to react to the force created by cornering. If we know what each end wants to do and can alter the suspension components (such as spring rates and roll center locations), then we can design our car to have a balanced setup prior to testing or racing. This is exactly what we shoot for by trial-and-error guessing. This method saves a lot of time and frustration.
Using a circle to represent...
Using a circle to represent the amount of weight on each tire, we can visualize weight distribution. With a low static distribution of weight (low crossweight percentage of 48-52 percent), after the weight transfer that occurs during cornering, an ideally balanced setup will have equal weight on the left-side tires and equal weight on the right-side tires.
With a higher static distribution...
With a higher static distribution of weight (high crossweight percentage of 56-62 percent), after the weight transfer that occurs during cornering, an ideally balanced setup will have equal weight on the right-front and left-rear pair of tires and equal weight on the left-front and right-rear pair of tires.