Improving the Fidelity of Structural Analysis and Sizing in Multi-Disciplinary Design of Space Systems

In modern conceptual/preliminary design of aerospace vehicles it is common for a large number of concepts and configurations to be rapidly explored. For each configuration, the structures discipline is responsible for determining an internal structural arrangement and detailed component design that minimizes mass while supporting external loads and other requirements. The proposed research presents a methodology suited for rapid design of structures which is capable of optimizing mass while easily meeting these requirements. Specifically, the methodology focuses on the stiffened panel optimization problem for metallic and composites. A change of variables is performed to allow accurate linearization of the design space, thereby greatly increasing optimization efficiency. The stiffened panel design space is recast in terms of equivalent smeared stiffness, using terms from the [ABD] stiffness matrix. This reformulation is enabled by the use of response surface equations to map the panel failure criteria (such as material failure, local buckling, etc.) to be a function of stiffness terms only. The resulting linear design space can be quickly optimized with the Simplex Algorithm (Figure 1). Thus, the approach is able to perform physics-based panel optimization with a level of efficiency appropriate for conceptual design studies. This approach is validated for a metallic and composite I-stiffened panel (Figure 2), as well as a composite laminate. Additionally, the methodology is demonstrated to couple well with the FEM-based design environment of a wing box for both metallic and composite construction. Overall, the methodology was shown to provide significant improvement in stiffened panel optimization efficiency over traditional tools while retaining accuracy within 10% of those tools.