Designs of Experiments (DOE) in Survivability Testing

(Source: http://jasp-online.org/asjournal/summer-2019/designs-of-experiments-doe-in-survivability-testing/)

(Source: http://jasp-online.org/asjournal/summer-2019/designs-of-experiments-doe-in-survivability-testing/)

July 16, 2019 | Source: JASP, Tom Johnson, John Haman, Heather Wojton, and Mark Couch,

The purpose of survivability testing is to provide decision-makers with relevant, credible evidence, conveyed with some degree of certainty or inferential weight, about the survivability of an aircraft. In developing an experiment to accomplish this goal, a test planner faces numerous questions: What critical issues are being addressed? What data are needed to address them? What test conditions should be varied? What is the most economical way of varying those conditions? How many test articles are needed? Using Design of Experiments (DOE) provides an analytical basis for test planning tradeoffs when answering these questions.

DOE is the process of determining purposeful, systematic changes to factors (the independent variables) to observe corresponding changes in the response (the dependent variable). For instance, an experiment may investigate the survivability of various aircraft against engagements from different air defense systems, as shown in Figure 1. The selection of factors, response, and design points is called the experimental design. DOE provides a framework for this process that accommodates a statistical model fit, allowing an analyst to extract cause-and-effect relationships about the system under test