Soft Walls or Soft Cars?

by Mark Cipolloni
February 26, 2001


What is rigidity?

Rigidity is generally thought of as how stiff an object or a structure is.  For example, a garden hose is not very stiff or rigid.  In fact it bends under its own weight.  However, a block of steel measuring 4" on all sides is very rigid.  Not only won't it bend under its own weight, it also bends very little when force is applied to it.

A materials measure of stiffness is called Modulus of Elasticity.  Steel, for example, has a very high modulus of Elasticity, whereas that garden hose, which is made of rubber, has a very low Modulus of Elasticity.

However, just because a material has a high Modulus of Elasticity does not mean it is 'stiff'.  A thin steel wire is not very stiff because it has a low Moment of Inertia.  The Moment of Inertia of an object is the sum of the products of each mass element of the object multiplied by the square of its distance from an axis (usually the center of gravity).  In layman's terms, if all the objects mass is located close to the center of gravity (think of the thin wire) it will have a low Moment of Inertia.

By moving the mass away from the center of gravity, you increase the objects Moment of Inertia.  As an example, a CART Champ Car chassis has a very high Moment of Inertia (both bending and torsional or twisting) because, as can be seen in the photo below, most of its mass is moved out to it extremities and the center is hollow.

Back side of a Champ Car 'tub'.  The 
hollow area is where the fuel cell resides.

Torsion, is twisting strain produced when a torque is applied to an object. For example, torsion is the strain experienced by a length of wire when a twisting force is applied to one end while the other end is fixed. Torsion can be measured by observing how much an object twists due to a given torque. For example, when a cylindrical object one unit long is twisted at one end, and the other end is held fixed, the amount the ends of the cylinder rotate relative to each other is a measure of the torsion. Engineering materials employed in rotating machine parts, such as engine crankshafts and ship-propeller shafts, must resist the torsional stresses induced by the twisting loads.  Whereas Moment of Inertia is a measure of an objects ability to resist bending, Polar Moment of Inertia is the measure of an objects ability to resist twisting.

A measure of an objects strength is directly related to both its Modulus of Elasticity and its Moment of Inertia or Polar Moment of Inertia.

Generally, long slender objects are not rigid, and short stocky objects are.  The hollow steel tubing used to make a Winston Cup frame is, by itself, not very rigid.  It won't bend under its own weight like the hollow tubular rubber garden hose will because steel has a much higher Modulus of Elasticity than rubber, but it will bend when relatively very little force is applied to it.

However, by making a space frame out of steel tubing, you can make a very strong and stiff structure.

Tubing made into a space frame 
protects drivers in violent crashes

Space frames are generally very stiff structures even though the individual elements that make up the frame, may themselves, not be very stiff.  That is exactly the case in a Winston Cup car steel space frame chassis

Since Dale Earnhardt's fatal accident last week at Daytona, safety in racing has received a lot of attention.  A lot has been written and discussed concerning the HANS Device because Dale Earnhardt died of a Basal Skull Fracture.  I firmly support the idea that drivers wear the HANS Device, however, there is another area of safety that needs to be addressed - Race Car Crumple Zones.  More specifically, should the concrete walls be made soft and crumple, or should sections of the cars be made soft and crumple.  That is the issue this article will attempt to address.

Any race car designer will tell you that the rigidity (see sidebar) of the chassis is a key component of a good handling race car.  A good chassis is one that offers good torsional rigidity as well as transverse and lateral rigidity.  A rigid chassis allows weight to be transferred from one end of the car to the other, from one side to the other, or diagonally from one corner to the other (Called cross-weight transfer).

The most important measure of torsional rigidity is how much stiffness there is from the center of gravity of the car to the front axle and from the center of gravity to the rear axle. This is part of the front weight transfer equation. If the distance from the center of gravity to the front axle is larger than to the rear axle you will have more front weight transfer (before springs and tires etc.).

ote various bars throughout a Winston Cup chassis space frame which gives it strength and torsional rigidly.

Unfortunately, in an accident, a very rigid structure does not crumple much and, therefore, much of its Kinetic Energy is lost in a very short timeframe and the object decelerates very rapidly. 

Crumple Zones
In recent years, the automotive industry has improved the crash worthiness of passenger cars by the use of Crumple Zones in the front and rear of the cars. Crumple Zones are created by the integration of variable grades of steel and composites into the front and rear-end assemblies of the automobile.  These Crumple Zones yield during impact, redirecting the energy of the collision, often reducing the chance of injury to the driver. 

This is best illustrated in the two examples below.  The first example shows a solid steel block hitting a cement wall.  The block does not crumple at all, stops almost instantly, and rebounds or bounces off with almost as much Kinetic Energy as it had just before it hit the wall.  In the second example a Coke can hits a wall.  However it crumples and deforms just as if you stepped on it, and much of its energy is dissipated over a period of time while the can crumples.

Solid steel block hits wall, stops almost immediately with very high 'G' force deceleration and rebounds.  In physics terms, this would be described as a very elastic impact. See this article for an explanation of Kinetic Energy.  

Aluminum Coke can hits wall, but decelerates much slower while it crumples.  Because it crumples and decelerates over a much longer period of time, it experiences very low 'G' force deceleration and rebounds with much less energy.  Energy is lost in the form of sound and heat.  In physics, this would be called a very inelastic behavior.

GM crash test.  The front of the car crumples, but the drivers compartment remains intact.

In an accident, the general goal is to keep the safety cell (where the driver sits) intact at the highest speed possible without killing the occupant, this implies that the safety cell has to be as stiff as possible (to avoid the collapse of the safety cell and the intrusion of the wheels, engine or steering wheel which can be deadly) and has to be surrounded by a crumple zone that is not too stiff (the sudden deceleration would kill the driver instantly) nor too soft (useless, the wall you'd hit would go right thru the crumple zone and you'd be killed anyway).  

A trade-off has, therefore, to be found for the strength of the crumple zone: at low speed (25mph for example), the car (car #1) with a very soft crumple zone might inflict less damage to the driver than a car with a stronger crumple zone: the driver would be shaken badly in the 2nd car, yet the driver "wouldn't feel anything" in the first car.  However, at 60 mph, the driver of the 1st car would probably be killed by the intrusion of the engine, the steering wheel, etc. where in the 2nd car, because of the " extra margin" allowed by the stronger crumple zone, the driver would probably be injured but still be alive.

A well designed Crumple Zone, one which extends the time over which a car decelerates,  exponentially reduces the force felt by the driver. 

It is, therefore, possible to find a car (# 1) that manages to protect the driver better than car # 2 at a speed of 25 mph, yet car #2 could be safer than car #1 at a higher speed than 60 mph because car #2 has a stronger crumple zone than car #1, allowing it a " last resort" extra safety margin that the driver would not have with car #1.

That is why it is crucial to see how well the structure of a car performs in a crash test.  Only the one whose structure hasn't started to collapse will allow the driver that extra safety margin in case of a stronger impact. Note that the size of the crumple zone is important : that is why, in general, minivans (with shorter "noses") don't perform as well as cars in frontal crash tests.  So crumple zones have to be "attached" to the stiffest cell possible and must not be too soft nor too strong! 

Crash tests results have nothing to do with the weight of the car tested.  It has to do with how well the safety cell-crumple zone-restraint system combo have been designed. So if you drive a very light car with good crash tests results into a wall, you will be better off in that car than in a truck, even 2 times heavier, that would have "bad" crash tests results.  However, if that truck hit the small car, even though that car has better crash test results than the truck, you'd be better off in the truck because of the weight difference.  In general, you'd be safer in a heavy vehicle hitting a lighter vehicle, regardless on how it performs in crash tests.

What happened to Dale Earnhardt?

Many people can't believe that Dale Earnhardt was killed at Daytona.  The accident just didn't look that severe.  Unfortunately I don't have any crash data to analyze because Winston Cup cars are not equipped with any CRASH boxes that are found on CART Champ Cars, F1 cars and IRL Indy cars.  NASCAR is taking a lot of heat because of this.  Therefore, it's hard for us to determine exactly what happened.  Yes, his seat belt may have broken, and yes, the HANS Device very well may have saved him.  However, the fact of the matter is that the vector component of speed that Earnhardt's car hit the turn 4 wall at Daytona was no more than 55 mph.  Stop for one moment and think about driving your car into an immovable wall at 55 mph.  I would not want to try it, even if it does have a crumple zone.

The one thing probably working against Earnhardt was the fact that the front of Winston Cup cars have become stiffer over the years in the search of better handling.  In the photo to the right (top), one can see that, when finished, the front assembly of a modern Winston Cup car is relatively stiff compared to the back (bottom photo to right).  That is one reason why drivers typically walk away unscathed in rear-end accidents - the backend of the car acts as a Crumple Zone, absorbing kinetic energy in the form of heat and sound as it folds up right to the back of the strong drivers compartment cage -  much like the Coke can did in the example above.  However, on front impacts, a modern Winston Cup car crumples up to a point near the front wheels, after which the chassis becomes very stiff, upon which deceleration becomes very rapid.

Picture the engine and transmission mounted in the chassis in the top photo.  Together they make for a relatively stiff structure.  Compare that to the relatively flimsy back of the chassis in the bottom photo. 

A properly designed front Crumple Zone on a Winston Cup car would probably mean the end of cars going back out on the race track after an accident just to collect points.  The front assembly would need to be designed to crumple in a more predictable manner than it does now.  That would probably render the car unfixable at the race track, or at least might require a nose change before re-entering the race.  I think it's better to make the cars safer in an attempt to save a drivers life, than to collect a few points trolling around the apron of the race track for three-quarters of the race.

Soft Cars vs. Soft Walls

Typically, the most lethal accidents are those in which a race car hits an immovable concrete wall and the 'g' forces felt by the human body are not survivable.  The crashes you see when a car goes somersaulting down the race track tend not to kill a driver because at no time does the driver decelerate too fast.  While the car is somersaulting, it's losing kinetic energy over a relatively long period of time (seconds vs. milliseconds when you hit a concrete wall).  Therefore, the race car industry has to take a serious look at the concrete wall impacts.

In order to lessen the blow to the driver, we are going to have to make a concerted effort to develop an acceptable 'soft' wall design, or we are going to have to design the race cars with Crumple Zones.  Both ideas have their problems:

Soft Wall Challenges on oval tracks

  • Cost to the track owners to outfit the many existing tracks

  • The variable resistance needed for light open wheel cars vs. heavy 'stock' cars

  • The potential for snagging a car because the wall will tend to pocket when hit.

  • The possible bounce back of a race car into traffic

  • The walls ability to perform a 2nd or 3rd hit in the same race.

  • After accident cleanup

Crumple Zone Challenges

  • Design and testing that will be required to make the front of the cars just stiff enough

  • Cost to the car owners to make changes to all their cars

  • The possible tradeoff between a stiff, good handling car, and one that is a little more flexible but gets around the race track a little slower..

  • The possible need to introduce space-age composite, energy absorbing materials, into the front of what is essentially a 1950's designed unsophisticated tube frame structure.

After studying this problem for awhile, I am of the opinion that without a doubt, soft walls work well for road courses and the inside walls of oval tracks.  They should be implemented immediately everywhere.  However, the outside walls of oval tracks present a much different problem, and the jury is still out as to whether a soft wall will work on outside walls.  The sanctioning bodies must step forward and start testing the various products out on the market today to determine if a soft wall will work.  It's too hard to model, only real live testing with properly instrumented cars will tell the whole story.

If I had to guess right now, I might lean toward making the crumple happen in the car rather than the wall.

What speed do you design for?

A lot of accident analysis has been done looking at the geometry of most oval tracks.  Unless a race car gets down into the grass infield, the steepest angle the car could hit the wall is 30 degrees (that is the angle of impact of the cg of the car with the wall). In Earnhardt's case his impact was probably more like a 20 degree hit.  The SIN of 30 degrees is 0.5.  That means a car traveling at 100mph would be the equivalent of hitting the wall at 50mph head-on.  The FIA standard test for soft walls is also about 50mph.  Data has shown that most impacts with the outside concrete wall are equivalent to a 50mph head-on impact, or less.  As I stated above, that's still pretty severe.

Compared to a 1950's tube frame stock car, a CART Champ car is made of space-age composite materials.  This view is looking down where the drivers legs extend.  The seat is removed.  The first bulkhead you see is about where the dash board is mounted.  There is another bulkhead further forward in front of the drivers feet.  They have about a 3' Crumple Zone in front of the drivers feet  The strength of that crumple zone is specified by CART to have a predictable crumple force. 

Using some of the equations for energy from our previous article on safety, one can determine what the size of a Crumple Zone should be.  Let's conservatively assume 75 mph to give ourselves some margin of error.

An object in motion has kinetic energy. The magnitude of the kinetic energy depends on both the mass and the speed of the object according to the equation:

E = 0.5mv2

where m is the mass of the object and v2 is its speed multiplied by itself.   For a 3500 pound car traveling at 75mph (110 fps), E = 0.5 x 3500/32.17 x 1102 = 658,222 ft-lb.


The change in a vehicles energy, ‘delta E’, can be derived from the equation:

‘delta E’ = (m x a)d

where ‘a’ is the acceleration (or deceleration, also known as negative acceleration) applied to the mass, ‘m’, and ‘d’ is the distance through which ‘a’ acts.  However, energy really does not enter into the analysis, but force, distance and time do.


Similarly, using Newton's second law, F = ma
(3500/32.17) x (60 x 32.17), one can determine that same car will apply a constant force of 210,000 lb. against the wall it is hitting until it came to rest (assuming a linear deceleration).  That force would be needed to design a properly stiff Crumple Zone.   The time, therefore, it will take that same car to decelerate so that the driver experiences no more than 60 g's is the change in velocity divided by the deceleration or, 110 fps/(60 x 32.17) = 0.057 sec., or 57 milliseconds.  That would be assuming that the Crumple Zone applied a constant resistance until the car stopped.  In reality the resistance is not linear and the instantaneous deceleration is usually about twice as great, hence the time of 57 milliseconds is only approximate and why we used 75 mph for our analysis when in fact most collisions with the outside wall are under 55 mph.


Assuming we don't want the driver to experience any more than 60 g's (60 x 32.17 feet per sec2), an effective crumple zone in any race car hitting a wall at 110 fps and decelerating linearly to zero, regardless of weight, would have to be about 3 feet long (110 fps/2 x 0.057 sec. = 3.135').  That's assuming a 75mph hit.  If we use a more practical number, about 55mph, the distance is 2.3'.  Another way to derive the same answer is to divide the Energy computed above (658,222 ft-lb) by the derived constant force of 210,000 lb = that same 3.13 feet.  Remember when your teacher tried to tell you that Energy = Force x Distance?  It's really true!


In round numbers then, a properly designed Crumple Zone for a 3500 lb. stock car would be about 2.3 to 3 feet long, be able to sustain a consistent force of 210,000 lb. for a period of at least 57 milliseconds, while decelerating the car from 75 mph (component of speed of the cg of the car into the wall).  Although the nose of today's Winston Cup cars are about 3' long before they crumple and the stiff part of the chassis is engaged, they crumple too easy and, therefore, don't keep the deceleration under 60 g's for a long enough period.  Drivers are also questioning NASCAR's decision last year to increase the strength of the front-end roll bars from 93-hundreths to 125-hundreths steel. The stronger front-end bars could make the stiff part of the car less crushable. And there are complaints about crews being allowed to use chrome-moly alloy in the bars; that makes for stronger bars, but chrome-moly doesn't bend as easily as other steel and can actually fracture. Plus if welding isn't done just right, the welds can crystallize.


The challenge to NASCAR will be to add enough resistive materials in the first 3' of the car so that it acts like a properly designed Crumple Zone for the weight of their cars.  And if that means using some form of honeycomb composite materials to do it, then so be it.


Racing will never be a 100% safe sport, but NASCAR certainly has a lot of room for improvement.  And they can start with the noses of their cars.


NASCAR, are you listening?