Fidelity Assessment of Spacecraft Structural Models Using the Spatial Nyquist Criterion

Abstract

This paper quantifies model fidelity for common spacecraft structural subsystems which predict dynamic displacement under operational conditions. In the preliminary and conceptual design phases, these models consist of finite elements such as beams, trusses, plates and concentrated masses. The fidelity metric quantified in this paper is based on the Nyquist criterion from signal processing and relates the fineness of the finite element mesh to the accuracy of an output figure of merit. Fidelity metrics for beam and plate models will be presented. It will be shown that using the fidelity metric provides a way of determining the suciently-fine finite element mesh required for a selected accuracy level a priori. This is presented as an alternative to model reduction and finite element adaptive mesh procedures. The fidelity of a nominal spacecraft telescope model is assessed and refined using the Nyquist fidelity method. A comparison of output figure of merit accuracy and computational time is made between the Nyquist fidelity method and the balanced model reduction method, and it is shown that using the Nyquist fidelity method saves 62% computational time over the model reduction method while retaining model accuracy.