Spacecraft multi-axis vibration testing: Model correlation, load synthesis and system identification

The launch vibration environment is inherently multi-axis, yet spacecraft ground testing predominantly relies on sequential single-axis excitation, leading to overtesting and extended campaign durations. Given the increasing industrial availability of multi-axis shakers, this article presents a comprehensive study on their viability for improving structural dynamic models. Firstly, it shows that for finite element (FE) model correlation, 3-axis (3D) shaker test data yields a higher frequency response assurance criterion (FRAC) for most sensors compared to the traditional single-axis approach, particularly in the low-frequency range. Since currently available electrodynamic shakers are often limited to 3 translational axes, while real launch-induced loads are fully 6-degree-of-freedom (6D), a bounded-variable least squares (BVLS) algorithm for deriving optimal multi-axis drive signals from arbitrary reference outputs is proposed. Applied here to several 6D-to-3D load reduction cases, the method successfully compensates for missing rotational degrees of freedom. Experimental validation using both unfiltered 3D BVLS and post-processed piecewise-linear specifications showed consistent matching of the 6D FE reference responses. Finally, for system identification, multi-axis test results are employed to directly estimate mass, stiffness, and damping matrices, capable of reproducing the structure’s dominant dynamics without a priori FE modelling. Overall, the findings confirm that a multi-axis testing framework provides clear benefits for both numerical modelling and ground testing of spacecraft structures, establishing 3D shakers as a viable alternative to 6D systems.