Fibre-reinforced polymer (FRP) composites are being used increasingly in turbomachinery components, due to their light weight and high specifi c strength. Bladed components are sensitive to vibrations, which are driven by the magnitude of excitation and their damping. Vibrations cause high-cycle-fatigue and eventual failure, so care must be taken to minimise vibration amplitude,
through engineered damping if possible. Modelling material damping in composites is
challenging due to their anisotropy but tailoring the layup of a laminate could potentially positively influence the material damping. This work presents a low-order modelling approach and experimental measurement technique, producing computationally efficient and simple damping
predictions for composite components with arbitrary geometry and layup.
This "smeared" approach, whilst similar to existing approaches, uses a strain energy computation to determine the material damping, but homogenises the effective properties of entire layups, rather than lamina properties, as typically used for macro-scale modelling. Initial experimental
validation of the approach showed it to predict damping well for abstract single-layup
specimens. Improved input damping parameters were produced through the development of a novel test rig, consisting of heavy tip masses attached to rectangular coupon specimens. This reduces extraneous damping contributions signifi cantly. The test rig facilitated further investigation into the scalability of smeared predictions, showing that the smeared elastic moduli and damping parameters can be used to represent the behaviour of thicker laminates effectively
if out-of-plane stress and strain contributions are accounted for during modal loss factor
computation. This investigation, coupled with input parameters gathered using the test
rig, provided the con fidence to apply the smeared technique to geometrically complex, multilaminates.
The technique predicted modal damping effectively, proven with full experimental
validation. Throughout the work, the smeared approach is shown to produce equivalent, and
sometimes superior, accuracy to the existing layered approach, with a signifi cant reduction in computational cost.