Wrapped Around the Axle

Wrapped Around the Axle

Balloon car designers Gene Wester and Chris Stell write:

In mid-November JPL announced an extracurricular pre-holiday fun contest among
employees to see how far a vehicle could travel powered only by two standard 9″
party balloons. We decided to collaborate, so we got together at lunchtime about
10 times over the past 4 weeks to work out design details and to test and
characterize the balloons. Chris did the bulk of fabrication at home. We tried
to keep costs low and spend money only where it produced a worthwhile benefit
(like ball bearings, instead of bushings). There are lots of design trades that
could be made, but we did not have time for much of that this year, so our design
was done mainly by instinct and modified by trial and error.

For those interested in design details, our car was constructed entirely of 0.032″
fiberglass sheet, with the exception of brass axles, steel bearings, and steel ballast.
It had two 5″ diameter front drive wheels, and one 3″ rear wheel. The 29″ long
body was a single beam with a hollow triangular cross-section formed by taping 3
strips together with mylar tape. At the front the triangular cross-section gradually
transitions to a rectangular cross-section with an open top for access to the front axle.
It looks somewhat like a slingshot dragster, but driving in reverse. The balloons
were tied together side-by-side and fit inside the hollow body; they were anchored
at the rear and connected at the front to a cord with a loop in the end that hooks
over a pin in the axle. The front cord (not the balloon) is wrapped around the 3/16″
diameter axle to stretch the balloon. Wrapping the balloon directly around the
axle would not have utilized the full balloon length to store energy. When the
balloon becomes fully contracted during a run, the looped cord slides off the
axle pin, allowing the axle to continue turning freely (without winding the string
in the opposite direction, which would have slowed the car).

Most of the key design concepts were defined fairly early: 3 wheels, drive wheels
in front, 3rd wheel in back, low friction, low frontal area, low mass composite frame.
We considered 3 ways to store energy in a balloon: twisting, stretching, and
inflating. Although we liked the idea of using air in the inflated balloons to
drive a piston, we thought an efficient piston/cylinder would be too difficult to
fabricate in the time available. We were originally going to build 2 cars–one
with stretched, and one with twisted balloons–but when we measured and compared
the potential energy available from twisting and stretching, we concluded that the
too much energy was lost with a twist approach.

We considered several beam shapes before we figured out how to put the balloons
inside a triangular cross-section. That concept not only minimizes bending forces,
but also results in a very stiff beam.

One early problem was how to attach to the ends of the balloons. We experimented
with several methods, and ended up using dacron lacing cord, tied with a
full-hitch and an extra half hitch. The lip at the open end of the balloon keeps
the knot from slipping off. We put a small ball inside the balloon to perform a
similar function at the closed end. The knots worked extremely well; during
failure testing the balloon failed before the knot slipped. We found that the
balloon became gummy and stuck to itself after a few stretches unless it was first
treated inside and out with corn starch.

We wanted to determine the optimal length to stretch the balloon in order to
recover the most energy during contraction, so we conducted some tests to measure
tensile force versus stretch distance. We discovered that force is a very nonlinear
function of stretch. We also learned that the energy recovered during contraction
is a small fraction of the energy required to stretch the balloon. Finally,
the balloon gets permanently weaker when it is stretched beyond a certain limit.
We tried to operate within that limit.

We cut out several wheel sizes (2, 3, 4, 5″) with the intent of running parametric
tests, but in the press of time we ended up using only the large size for drive.
One of the things we would change would be to use even larger wheels to step up
the effective gear ratio.

The rear wheel steering/suspension went through 3 major revisions before we found
something that worked well. The rear wheel was initially attached to a bent
wire with a vertical axis of rotation, which resulted in directional instability.
Angling the axis of rotation helped, but did not solve the problem. Eventually
we attached the rear wheel to a horizontal fiberglass yoke that could be
precisely aligned with a slotted adjustment; the thin fiberglass strip also
acts like a shock absorber in the vertical direction.

Our initial objectives were to minimize mass, friction, and moment of inertia,
but our vehicle was so light that the wheels spun. The traction problem was
solved by a combination of increased friction and increased mass. The
coefficient of friction was increased by means of a rubber band stretched around
the rim of each wheel (and it looked sharp too). Adding mass near the drive
wheels not only increased traction, but also reduced acceleration, top speed,
and associated aerodynamic drag. The final weight was less than one pound.

Our car won the grand prize for distance. On the official run it traveled 71′
before it went out of bounds and crashed into a planter. After the contest
ended several entries did a second run, and our car went 97′ down the center of
the course. Our car was also selected as “Most Artistic” by judges from Uncle
Milton Industries and Mattel Corporation.

Overall, we had a lot of fun and even learned some things along the way.
There is already talk of another design contest next year that would be
totally different. I think we are hooked!

MB 1/9/99