Semi-autonomous analysis of large aircraft Ground Vibration Tests

Abstract

The ground vibration test (GVT) is performed only weeks before the maiden flight. Therefore, the measurement campaign and data analysis are highly time critical. Nevertheless, the results must be highly accurate, because they are used to perform the critical flutter calculations. In this work, a semi-autonomous system is presented that exploits few analysis results of an engineer and learns optimal ways to perform the analysis of the GVT data depending on the excitation conditions. After the system is trained for several excitation conditions, it analyzes new data sets fully autonomously within seconds with high accuracy. The system was tested onsite at the GVT of a research aircraft and the results are compared with the state-of-the-art data analysis for GVT. The semi-autonomous system is capable to find all modes of the structure with the same level of quality as the state-of-the-art analysis chain. However, in significantly less analysis time. Additionally, the results are userindependent and fully reproducible. The system is promising to improve and standardize the procedure for vibration test data processing and identification of modal parameters. Since the systems learns the way to analyze the data best based on the engineer's results, it can be applied in a broad field of applications in research and in industry.