Why Our Chassis Looks The Way It Does
Any good chassis must do several things:
  1. Be structurally sound in every way over the expected life of the vehicle and beyond.  This means nothing will ever break under normal conditions.
  2. Maintain the suspension mounting locations so that handling is safe and consistent under high cornering and bump loads.
  3. Support the body panels and other passenger components so that everything feels solid and has a long, reliable life..
  4. Protect the occupants from external intrusion.

In the real world, few chassis designs will not meet the criteria of #1.  Major structural failures, even in kit cars, are rare.  (Here's an exception.)  Most kit designers, even if they're not engineers, will overbuild naturally.  The penalties for being wrong here are too great.  The trouble is, some think that having a "strong" (no structural failures) chassis is enough.  It isn't.  Read this article from the July 1999 Machine Design magazine!

Structural stiffness is the basis of what you feel at the seat of your pants.  It defines how a car handles, body integrity, and the overall feel of the car.  Chassis stiffness is what separates a great car to drive from what is merely OK.  

Contrary to some pronouncements, there is no such thing as a chassis that doesn't flex, but some are much stiffer than others.  The range of chassis stiffness has varied greatly over the years from about 500 lbft/degree in a Morgan to more than 20,000 lbft/deg in a modern race car.  The ERA 427's chassis runs about 3500 lbft/degree.  Not high by current sedan standards but about as high as you can get in a roadster whose body sits mostly on top of the chassis.

Different basic chassis designs each have their own strengths and weaknesses.  Every chassis is a compromise between weight, component size, complexity, vehicle intent, and ultimate cost.  And even within a basic design method, strength and stiffness can vary significantly, depending on the details.  There is no such thing as the ultimate method of construction for every car, because each car presents a different set of problems.  Below, I have summarized the characteristics of some chassis alternatives.  Remember, though, that detail execution is as important as the basic design, if not more!

Some think an aluminum chassis is the path to the lightest design, but this is not necessarily true.  Aluminum is more flexible than steel.  In fact, the ratio of stiffness to weight is almost identical to steel, so an aluminum chassis must weigh the same as a steel one to achieve the same stiffness.  Aluminum has an advantage only where there are very thin sections where buckling is possible - but that's not generally the case with tubing - only very thin sheet.  And even then, aircraft use honeycomb'd aluminum to prevent buckling.  In addition, an aircraft's limitation is not stiffness, but resistance to failure.


Backbone: The tunnel becomes a primary load bearing member. This is a potentially fine design, and if we were building a new car from scratch, we would seriously consider a backbone.  But , this is not a new car, it's a replica of a classic!  Because it is designed around the original Ford engines (and we wanted our customers to have several different transmission choices), the bulk of a compatible structural tunnel was unacceptable, especially considering the passenger compartment was a fairly narrow one to begin with. A backbone would make it impossible to maintain the look of the original interior and engine compartment.  It would also create servicing difficulties.

A variation to the sheet metal backbone is one that uses small tubes to create the central structure.  TVR's Griffith was built like that - with an enormous tunnel.  The Shelby Daytona Coupe added a tubular backbone to the original 289 chassis. It probably added 50% to the overall stiffness of the car!  See below.

 Then there is the issue of engine compartment esthetics.  With our rectangular tube chassis, we can duplicate the round-tube X (with the 427SC) or the spring tower (with the 289FIA) at the front of the engine to maintain visual accuracy.


Space frame: A true space frame has small tubes that are only in tension or compression - and has no bending or twisting loads in those tubes. That means that each load-bearing point must be supported in three dimensions.  It is nearly impossible to build an efficient space frame around the Cobra body. The rockers are simply too shallow, and the tunnel shaped incorrectly to make a reasonably triangulated structure.

Remember the 300SLR Mercedes?

(shown at the right)  It had rockers 12 inches tall and 10 inches wide and the chassis used hundreds of separate tubes. It was difficult to build and a nightmare to fix. The "space frame" chassis that is currently built for another replica simply uses smaller tubes, many carrying bending and torsional loads.  It may look impressive, but functionally it's a bad compromise.  Simply more complication without improvement.  
Space frame

Mercedes 300SLR

Consider - the bending stiffness of a tube increases the by the square of the diameter of the (equal-wall-thickness) tube, and the torsional stiffness by the cube of the diameter, while the weight goes up linearly.  The bottom line is - sometimes you're better off with a large tube.


Monocoque

1958 Lotus Elite

Monocoque:  An airplane (with a stressed outside skin) is close to a true monocoque.  In the automotive world, it's time to compromise again, but the street car that compromises the least is probably the 1958 Lotus Elite.  The design was  made possible by the use of large fiberglass panels - otherwise the tooling and construction costs would have been tremendous. In the real world, the interior panels are stressed, but many cars have an aerodynamic facade of 'glass or aluminum.  

The original GT40 - and our ERA GT - have a semi-monocoque chassis.  The heaviest (steel) main panel on our ERA GT is only .045" thick, and most panels are only .032"!  Reinforcements are required at the suspension points where there are local high loads.  With the rockers 10" high x 9" wide, the net result is an incredibly stiff structure.  But you can't build a classic roadster like this.

Monocoque chassis
ERA GT Chassis

"Ladder" frame:  The ladder frame is a shorthand description of a twin-rail chassis, typically made from round or rectangular tubing or channel.  It can use straight or curved members, connected by two or more crossmembers.  Body mounts are usually integral outriggers from the main rails, and suspension points can be well or poorly integrated into the basic design.  The original Shelby 289 Cobra used 3" round tubes, a very flexible design that worked with stiff transverse-leaf springs for adequate but primative handling.  The 427 was updated to 4" round tubes to allow the more modern suspension to  work properly.  Both chassis were very simple to build - and good enough for their time.

Real chassis

An Original 289 chassis


The Semi-Backbone

The Shelby Daytona used a modified 289 chassis made into a tubular semi-backbone design to correct the extreme flex of the original design.  You can see how it looks by visiting the site of someone ambitious enough to try to build a Daytona from scratch!

Daytona chassis
Daytona Chassis


The ERA chassis uses 4" x 3" x .125"W structural tubing in a complex design meant to take suspension and body loads efficiently, while maintaining the original look from outside and in the engine compartment, simutaneously allowing easy service and assembly.

Roll bar, body and door mounting points are built into the basic design for maximum efficiency.  There are 4 crossmembers plus an "X" member for maximum torsional stiffness.  We even box in the "X" for extra strength!

Yes - this chassis is somewhat heavier than most ladder designs, but it is also by far the stiffest.  A compromise that no ERA owner regrets!

ERA 427 Chassis
ERA Chassis (FIA similar)


Round vs. Rectangular frame rails: There has been a lot tossed around regarding whose chassis - and what kind of tubing - is "strongest."  Factory Five is numerically the biggest exponent of round tubes, but many others have preceded them.  We chose to use rectangular tubing in our chassis for several reasons: Under pure vertical bending load, 4" x 3" rectangular tubing is about 37% stiffer than an equal wall thickness 4" round tube. This is especially important because a roadster doesn't have a roof to stiffen the passenger compartment. Not only can you feel a lack of "solidness" with a flexible chassis.  Your variable door gaps will also make latching unstable - and even ocassionally cause paint chipping as the doors meet the main body!

You can see below that transverse members have little effect on beam stiffness.  You just add up the individual stiffnesses of the components.  We also have an "X" member, acting as an additional longitudinal beam reinforcement and as two transverse members.  A round tube chassis is extremely difficult to "X" brace.

Twist/Bend/Shout

A little light on Torsional Stiffness

Even though an individual rectangular tube is about 2% less stiff in torsion than the equivalent round tube, we must consider the chassis design as a whole. For each transverse tie-in we create a system that becomes more like a single large tube spanning the whole width of the chassis- the ultimate in efficiency. We have integrated 7 transverse members along our main rails in such a way that the chassis has much more torsional stiffness than the tubes taken individually.  We even put extra braces on our central "X" member to make it even stronger.
The stiffness of an ideal unitized structure is proportional to the square of the distance of the components from the centerline. Double the distance and you have four times the overall stiffness. While practical automotive considerations eliminate an ideal connection between the rails, widely spaced tubes that are tied together well work more efficiently than the same tubes on a narrower base. The original 427 Cobras' rails were only 20 inches apart. Ours are spaced at 27 inches on center through the middle of the chassis, one of the widest spacing in the industry. And we still are one of the few in the industry that have left room for an undercar exhaust outside the rails.

THE BOTTOM LINE:

The E.R.A. chassis is one of the strongest and stiffest of the industry.  
And the difference is
easy to feel on the road!

Please use your browser's <Back> to return

Avoiding bad designs From Machine Design Magazine, 11/20/1997
A few basic guidelines for preventing troublesome designs include:
  • Avoid complexity wherever possible. Strive for lower rather than higher component densities and use multifunctional components that reduce the total number of components needed for any given assembly.
  • Minimize variety, such as the number of different types of component or values.
  • Shun attributes known to have caused problems on previous projects. And keep an active database of new problems as they occur .
  • Know the limits of your company's production capabilities and design in margins for error. It is not always necessary to have an exact knowledge of the process capability. As a matter of fact, it is often best to underestimate the process' capabilities (i.e., assume more restrictive process tolerances than actually exist). This increases the probability the process will accommodate the design.
  • Seek innovative ways to prevent assembly errors.
  • Document every success; and failure as a design changes. Incorporate these experiences into formal design guidelines.

  • Make accurate prototypes of new designs and test them rigorously prior to approval for production. Prototypes should certainly go through harsh environmental stress screening.
  • Maintain formal multidepartmental design review boards to assess all new designs. Restrict turnover of participants in this review process since the ability to identify potential problems in the design stage can be acquired only through considerable experience.
  • Subject all new designs to evaluations, such as failure mode and effects analysis (FMEA), by the design review board as well as the design department. FMEA need not be complex. Determine the consequences of failure for each component and invest the most resources in techniques that minimize the possibility of the most serious failures
  • Employ formal CAD1ibraries that are constantly updated in accordance with plant capabilities rather than arbitrary industry "standards."
  • Stay constantly aware that the costs of initial design mistakes have never been greater and are constantly increasing.