A Light Goes Off

I was attending the PRI show in Orlando this past fall and I heard about some new type of spring that my old friend Kelly Falls and his engineer, Mark Campbell, had on display. It's a carbon-fiber disc bellows spring that is designed to replace a steel spring in certain applications.

One could see this being used and highly appreciated in IndyCar and Formula 1 racing and other formula-type of cars because they're lighter and take up less space. The jury is still out on whether it will be useful for short track racers due mainly to the limited travel before spring "coil" bind, but this item is still very new and development is still underway.

I saw and envisioned something different. I had an idea. For some time I have thought that if we could get the car down on a bump apparatus that could maintain a more constant spring rate throughout its useful range of travel, we could then predict and manage the setup.

If, say, we know the spring rate of the bump was 1,500 ppi through an inch or so of travel, then we could design the setups and balance the two ends of the car accurately. When I returned home, I envisioned how to install those bellows springs on a Late Model to test the theory.

The car would have to be a big spring design so we could use the shock for placement of the bellows springs without being impeded by the coilover spring. The idea was to install light big springs for static ride, then allow the shocks, with the bellows springs installed on them, to take the load in the turns. If we used a shock that would hold the front end on those bellows springs, even better.

All I needed was a set of bellows springs with a 1,500 ppi rate, an aluminum sleeve to take up the gap between the bottom of the shock and the bellows and a few other items to make this work. For a car and team, I had an ace in the hole.

Dick Anderson designed the Anderson Elite chassis that Port City Race cars sells to this day and of course Dick runs one of these chassis. It's a big spring design, much like the stock version front ends. He agreed to work with me to test this approach and we selected Orlando Speedworld for two reasons. One, it was close by; and two, it was more of a handling track with lower banking and shorter straights than a track like New Smyrna, another track that was close by.

Our driver would be Dalton Zehr, a very capable shoe who won more than a few races up north driving Gene Coleman's cars this past summer. He has also been our test driver for our G.R.E.E.N. Racing project Camaro as well as other projects including the Late Model stock car we took to North Carolina a few years ago.

The Setup

So, to configure all of this, I had a couple of 2.5-inch inside diameter aluminum tubes made up in an 8.00-inch length and I took them to a local Grand Am race shop where Stan Flis machined the ends perfectly square. This was important in that these tubes would be in contact with, and needed to be square to, the bellows springs.

We mounted the bellows springs onto double adjustable shocks that had been specifically designed and valved for bumpstop use where the rebound at 3.0 inches per second was around 1,050 pounds for both shocks and in the LF shock, the piston had no bleed hole and the bleed adjuster was almost closed off for this test, cracking the bleed about a half sweep. The compression in both shocks was around 100 pounds. at 3.0 ips.

I developed two setups using software. The first one was in harmony with an article I wrote that described how to have a zero roll angle. The idea was to design the front so that it rolled to a positive number, something it would always do anyhow, and to cause the rear to roll to a negative, but equal number related to the front.

In this case, the front roll angle was 0.70 degrees positive and the rear was 0.65 degrees negative. The net roll was near zero. To accomplish that, we installed a Left Rear 165-ppi spring and in the RR we installed a 410-ppi spring. The Panhard bar was at 10.5 inches on the left and 11.0 inches on the right measured from the ground.

The other setup I developed caused equal roll angles, both positive, in the front and rear. The thinking was that if the zero roll design did not work, and I'd really never tested it before, I could go with the second setup. It only involved switching to a 275-ppi RR spring with no change to the Panhard bar height.

The software recommended a crossweight distribution of 53.0 percent for this car, but Dick always ran 56 percent in his cars, so we went with Dick's cross. In setting the cross, we first set it with the car at ride height with it supported by the big springs.

Then we removed the big springs and let the car sit on the bellows springs. We then adjusted the rings on the shocks to attain a 56-percent crossweight while on the bellows springs and left the rear springs like they were at ride height. This meant that the car would experience no load distribution change on the tires going from riding on the big springs to riding on the bellows springs.

Incidentally, when we initially dropped the car down onto the bellows springs, we had approximately 1/16-inch clearance for the front valance. So, we adjusted the bellows heights up so that we had about a half inch of clearance.

While the car was down on the bellows springs, we also changed the front cambers to what we thought would work and knew we might need to make adjustments at the track. We set negative 5.0 on the RF and positive 2.5 on the LF tire. We also installed a medium 13/8-inch antisway bar.