Advanced non-linear models for the design of deployable composite space structures

Future space missions aim to bring back robotic devices and astronauts to the moon and later to Mars. Such challenging goals require overcoming the actual state of the art in many engineering fields and developing innovative solutions able to fulfill the mission requirements. The lightness and compactness of the spacecraft have always been a key aspect of space applications since the volume and payload offered by the launchers are limited. A promising technology that offers a huge reduction in the volume of large space structures is the use of deployable/inflatable structures. These structures can be folded in a small volume during the launch and then can be deployed during their use in space. Even though some demonstrators or small applications have been used in the last years, this technology is still far to be considered reliable to be applied in primary structural components. The design of these structures, especially when they are built using composite material, is very challenging since common numerical tools fail in the prediction of their response. The ineffectiveness of classical analytical approaches can be blamed on the laminated nature of these structures, the extremely low thickness, and the highly non-linear response. The present work proposes the use of advanced numerical models, derived in the framework of the Carrera Unified Formulation, to design composite boom typically used in large space structures such as antenna or truss structures. The present models, one- and two-dimensional, take advantage of an enriched kinematic approximation that led to accurate results even in the case of laminated structures. Equivalent single layer and Layer Wise models of different orders have been considered exploiting the convenient formalism offered by the Carrera Unified Formulation. A total Lagrangian formulation has been used to derive an efficient nonlinear model able to predict large displacements of composite structures. Different geometries and materials have been considered and the results have been compared with those from literature. The results show the great advantages offered by the use of higher-order models since they can provide accurate three-dimensional solutions with a fraction of the computational cost required using a classical approach.