Inverse Multi-Scale Robust Design of Composite Structures Using Design Capability Indices

Abstract

Composite materials are heterogeneous materials, and are hierarchical in nature consisting of multiple length scales. In the design of structures with composite materials, the micro-structure of the materials have a direct bearing on the final behavior of the structure. The deviations in the bulk material properties are caused due to uncertainties associated with the micro-structures and its propagation through different length scales. Uncertainties in the design parameters (geometry and materials properties etc.) at macro-scale also contribute to variations in the final behavior. Currently, these uncertainties are included as a large factor of safety in deterministic design, which may result in over design of the product. The robust performance of the structure can be achieved by considering these uncertainties explicitly in the design process. In this paper, a method for designing a robust composite structure subjected to different loading conditions is illustrated. Structural models are run to compute robust material properties and geometries for different load scenarios that yield most robust materials and micro-structures. Most robust combination of material and geometries is selected that results in most robust performance under all loading scenarios. These materials are designed using multiscale models in which micro-structural uncertainties are accounted. The uncertainties in the material properties and geometrical parameters at different length scales are explicitly modelled as ranges in the set of input parameters. Final performance variations are calculated using design capability index. Consolidated single material parameters and dimensions are selected using efficiency metrics. Design capability indices are formed as goals and constraints in compromise decision support problem. Robust micro-structures are designed inductively rather than deductively.