Applying experimental design in a debris dispersion model study

In the future, it is envisioned that daily space launch and return operations will occur at "spaceports" throughout the United States and the world. A projected increase in launch/return rates, and growing capacity demands on the National Airspace System, will require more efficient integration of air traffic near spaceports. This research addresses the debris dispersion hazard, which is one of the major safety issues associated with space launch/return vehicles and aircraft operating in shared airspace.;The Common Real-Time Footprint (CRTF) program, which is currently used by U.S. Eastern and Western Range operators, predicts the debris impact area and associated risks for a given set of launch parameters. The debris hazard modeled by CRTF can affect the size of the restricted airspace surrounding a launch site.;The CRTF program uses launch vehicle state vector, wind, the aerodynamic characteristics of the fragments, and other factors in modeling the debris dispersion. In this research, experimental design and statistical simulation techniques are used in a study of six key factors that affect the debris impact area. The specific event that was studied is a simulated catastrophic failure of Space Shuttle (STS-75) and Delta IV vehicles during a nominal ascent trajectory.;A two-level full factorial experimental design requiring 64 (2 6) independent runs was used to evaluate the six factors, and all possible combination of factors were evaluated for main effects and interaction effects. The results were evaluated for statistical significance using analysis of variance (ANOVA) techniques. Results showed that both the Shuttle and Delta IV produced three common main effect factors: wind (A); ballistic coefficient (C); and altitude (F). The Shuttle analysis identified three two-factor interaction effects: AC; AF; and CF, and the Delta IV analysis had only two two-factor interactions: AC and AF. Since both vehicles' interaction effects were a combination of only three main factors of the six variables studied, this result suggests that researchers only need to focus on these three factors in similar studies. A two-level fractional factorial design of resolution 6 also is shown to be a viable method to achieve statistically significant results and acceptable predictions.;First order prediction equations were developed for the sets of three factors that produce significant effects. An analysis of the residuals from the first order prediction equation suggested that second order influences existed, so a second order analysis was performed. The adjusted coefficient of determination (Ra2) from the second order model was slightly better than from the linear equation. However, both Ra 2 calculations produced acceptable values close to 1.0, and reasonable predictions could be obtained from both first order and second order analyses.;Studies targeting other operational scenarios could be performed to further investigate airspace requirements and range safety procedures. This type of research could also be applied to developing a common automated decision support tool that supports both air traffic and space launch/return. These results, and other related applications of experimental design and statistical simulation techniques, are detailed in the conclusions.