Discrete Material and Thickness Optimization of laminated composites using aggregated high-cycle fatigue constraints

This work presents an efficient approach for gradient-based structural optimization of multi-material laminated composites where high-cycle fatigue constraints are included by a novel approach. The parameterization of the structure is done by the Discrete Material and Thickness Optimization approach, such that an optimized combination of material, fiber orientation, layup sequence, and layer thickness can be obtained. The high-cycle fatigue analysis approach used takes offset in methods typically applied in the wind turbine industry for blade design. The variable-amplitude loading is quantified by rainflow counting assuming proportional loading. A constant life diagram approach is used to calculate equivalent stresses from the amplitude and mean components, taking into account mean stress of various magnitudes by interpolating between their respective S-N curves. Reversals to failure are computed from the SN curves, and damages are summed using cumulative methods such as the linear Palmgren- Miner damage rule. The high number of local fatigue constraints is reduced by use of aggregation functions. This also enables efficient design sensitivity analysis using the adjoint method. The potential of the optimization approach will be demonstrated by a number of examples.