THE ART OF TESTING - Derby Tech - February, 1985

by Tex Finsterwald

In gravity racing, as in most endeavors in life, the people who work the hardest are usually the most successful. In developing a gravity racer, the hard work starts with thought and research about how and what to build and continues through preparation for race day with no letup in between. One important, but often neglected facet of producing a quick racer is testing. Unfortunately, testing turns out to be a lot like work, and more often than not fails to produce the desired increase in speed.

Probably a major factor in the success of competitors from the greater Flint area over the past dozen years has been that we all do considerable testing. When we go racing, we are confident that we will be running our best setup. Over the years, I've seen many people come to race and say they are trying a new setup. These people frequently say they have made an improvement based on their performance against another competitor, or how they finish in the results. What they fail to realize is that the competitor has been working and changing also; and that he is likely running something different from what he ran at the last race. To be confident in evaluating setups, You must compare setups to a common benchmark.

One common method of testing is coast testing or rolling for a distance. In this method, the racer is allowed to accelerate downhill and then roll to a stop either on level ground or on an uphill slope. The exact spot where the racer stops is marked. In coast testing, several warm up runs should be made prior to actual testing. The starting position on the track or street should be located so that it may be repeated exactly on each trial. The surface temperature of the track should be monitored to observe if it changes during the test session, because it can affect the test results. Also, the direction and speed of the wind should be watched. An ideal day for testing is a calm one, and it is best to test over a time span when the temperature is constant. Assuming that one has tolerable conditions, several trials should be made using a given setup. It is not uncommon to find considerable variations using the same setup; although, sometimes the results are amazingly repeatable. Once a benchmark has been established, the setup on the racer is changed, and the tests are repeated. If the car goes farther on the average with the new setup, You can feel it will be a quicker one. Many different setups can be compared in this manner. It is good practice to repeat trials of trip setup which appear to be best to insure that the results are repeatable. Coast testing is best suited to concepts that do not affect the aerodynamics of the racer because it may not be carried out at the same speed as those encountered in actual race conditions. Coast testing can be done with only three people; a starter, a driver, and a catcher.

The method of testing performed by the writer is to run the subject car against a test car which is held constant during the testing. This method is more involved because it requires a suitable test track, a consistent starting method, a method of judging relationships between two racers at the finish line (camera or timer), two racers and good drivers, and a number of people at the starting area, the finish line, and the run out area. In one test session, we had ten people involved, and that was none too many. In this type of testing, one car is held constant during the tests and the other is changed to try out the items being tested. We make four runs with a given setup so that each car runs in each lane with each set of wheels, and then determine an average difference between the racers. This is sort of a double timer or photo swap. We feel this helps to compensate for changing track conditions over a long day.

In both types of testing, it is necessary to plan ahead as to what will be conducted in order to make good use of testing time. A demanding day of racer against racer might be spread over eight to ten hours, and thirty two runs would be an ambitious number of runs, considering the time required to make changes. Thirty-two runs may seem like a lot, but at four runs per setup it an amounts to eight setups tested. When one considers that certain setups may need to be retested, it becomes apparent that thorough testing is truly time consuming.

A pitfall to be avoided in testing is mistaking the cause of differences in setups with conditions other than those being tested. An example of this might be the rear axle of the constant car working loose and producing results which show the wrong setup being best. Also, it pays to be open minded so that preconceived ideas don't blind one to what is happening in testing.

In spite of the time and effort required., I maintain that good testing is the way to find out what's going on. If you don't test it, you're just guessing.

Return to the Derby Tech Page