Application of Design of Experiments to Flight Test: a Case Study

Aaron A. Tucker
Gregory T. Hutto
Cihan H. Dagli, Missouri University of Science and Technology

This document has been relocated to http://scholarsmine.mst.edu/engman_syseng_facwork/206

There were 33 downloads as of 27 Jun 2016.

Abstract

Modern flight test tends to be a complex, expensive undertaking so any increases in efficiency would result in considerable savings. Design of experiments is a statistical methodology which enables a highly efficient investigation where only the samples needed are collected and analyzed. the application of design of experiments to the design of flight test can result in a significant increase in test efficiency. Increased information is garnered from the data collected while the number of data points required to understand the system is reduced. in this effort, an actual flight test program serves as a case study to compare and contrast five different designs to explore the flight test envelope: the classic subject-matter-expert (SME) generated survey method, the SME-generated points augmented to a relatively fine mesh orthogonal analysis of variance design, an axial central composite design (CCD), a face-centered CCD plus simplex design, and a Simpson-Landman embedded face-centered CCD. the axial CCD is further expanded by a single point to illustrate the flexibility of the design in response to the interests of the test team. the case study data are analyzed using each designed experiment, and the results are compared and contrasted as a cost-benefit relationship between flight test resources expended (i.e. flight hours) and system understanding gained (i.e. statistical confidence and power). the design of experiments methodologies, as applied to this case study, generally show a 50 to 80 percent reduction in flight test resources expended to gain similar levels of understanding of the system under test. These savings can be applied to other programs, used to educate design changes before testing an improved system, allow for flexible investigation into areas of interest to the test team, or replicate the test points resulting in a better understanding of systemic error. in an era of restricted budgets and timelines, careful design and thoughtful analysis of flight test experiments can make the difference between a failed or cancelled flight test program and the successful fielding of a needed capability.