Skyward has been applying Design of Experiments (DOE) statistical analysis in areas which had not previously applied such rigorous experimental methods. The application of DOE methods improves test and analysis program planning, produces higher quality test data, and enables greater confidence in the conclusions generated from testing. In working with the U.S. Air Force, Skyward has been implementing DOE methods in ballistic live fire testing and in pretest predictions derived from experimental results. This testing has involved analysis of ballistic munition functioning and fire ignition and sustainment potential. The use of DOE and advancements in high-speed video capability has expanded the investigation of this complex phenomenon. In the KC-46 Live Fire Test & Evaluation (LFT&E) program, Skyward has applied statistical analysis throughout dry bay fire vulnerability testing and used virtual testing from LS-DYNA analysis to downselect test areas and shotlines in hydrodynamic ram testing of wing structure.
Skyward personnel have graduate level training, including the Design and Analysis of Experiments course at the University of Dayton and have achieved certificates for DOE 0, DOE I, and DOE II training through the Air Force 96th Test Group. Skyward personnel are proficient with Minitab, Design Expert, and have experience with JMP. These personnel are proficient with conducting 2k Factorial designs and experiments, Analysis of Variance (ANOVA) methods, and have examined the application of non-standard analyses such as Analysis of Co-Variance (ANCOVA) split-plot designs, Multivariate Analysis of Variance (MANOVA), etc.
Skyward uses DOE principles and practices to develop test matrices during testing planning, optimizing the required number of tests to save costs, while ensuring the factors examined are most likely to provide the greatest information obtainable. Skyward is using DOE analysis of previous tests to perform pretest predictions for future testing and to develop inputs for computer models. Applying DOE methods to test results, enables Skyward to maximize the knowledge gained from the testing and confidence in the conclusions. It also enables the validation of statistical models and highlights areas requiring improvement. The updated model is then, in turn, utilized with higher confidence over a wider area of input parameters and the combination of model and test data provides a better overall picture of the evaluated system’s response. Skyward also uses statistical analysis as a quality control process, ensuring that data outliers can be thoroughly explored. This often leads to improvements in test methodologies or the recognition of test areas requiring further exploration.