A Statistical Method Useful in the Development of Jet Engine Fuel Controls

Since most jet engine fuel control development programs require working to tight schedules, changing engine requirements, and in new areas of technology, the distinction between development and production becomes less apparent and the two activities, for a time, are carried on simultaneously. A development test program must take this into consideration.

Many problems encountered during the development period are straightforward and will yield quickly to the technical knowledge and experience of the engineer. When they do not yield to this approach, other more powerful statistical analytical techniques may be required.

One such technique is “Random Balance.” It is an experimental design which permits simultaneous study of the effect of many inputs with little added cost or delay in the program. It is a screening experiment that isolates the inputs that have the greatest effect on outputs under study. It assists in sorting the vital few from the trivial many. Furthermore, the technique does not replace the classical statistical experiment, but directs it to those areas where rigorous analysis is required.