Reliability of Experimental Modal Data Determined on Large Spaceflight Structures

Abstract

Launcher components, large payloads – and particularly in the past – also all major Space Shuttle payloads have required modal survey tests to determine the modal parameters with high accuracy and reliability to validate the respective dynamic structural models for coupled loads analysis and flight piloting, but also for system level structural qualification. The modal survey test methods applied usually consist of the classical ground vibration test using tuned sine excitation and the modal analysis of measured frequency response functions. While the tuned sine excitation directly results in well understood and established modal parameters – typically of the fundamental or primary modes, the modal analysis is used to complement the modal model with respect to the secondary modes. This two-folded approach serves both for high accuracy of the modal model considering the fundamental modes as well as to minimize test effort to complement the modal model in the higher frequency bands. Almost all available modal analysis tools – besides of very specific ones – which are available today are based on linear models and on a more or less statistical approach to estimate the modal parameters. With respect to real measured transfer functions, in particular when modal density increases and some non-linearity is present, which is typical of large space structures, the modal analysis results inherently suffer from a considerable amount of scatter or uncertainty. In this paper, application cases are studied to provide insight in the actual challenges to extract reliable modal parameters from large spaceflight components. Based on this, a consolidated strategy is discussed to extract the best match between experimental data and experimental modal model. In addition, the remaining uncertainty may be quantified as well.